0%(1)Il 0% ha trovato utile questo documento (1 voto)

97 visualizzazioni372 pagineSafety of Hystorical Stone Arch Bridges

Jan 31, 2019

© © All Rights Reserved

Safety of Hystorical Stone Arch Bridges

© All Rights Reserved

0%(1)Il 0% ha trovato utile questo documento (1 voto)

97 visualizzazioni372 pagineSafety of Hystorical Stone Arch Bridges

© All Rights Reserved

Sei sulla pagina 1di 372

Arch Bridges

123

Dr.-Ing. habil. Dirk Proske, MSc. Dr.-Ir. Pieter van Gelder

University of Natural Resources Delft University of Technology

and Applied Life Sciences, Vienna Section of Hydraulic Engineering

Institute of Mountain Risk Engineering Stevinweg 1

Peter-Jordan-Street 82 2628 CN Delft

1190 Vienna The Netherlands

Austria p.h.a.j.m.vangelder@tudelft.nl

dirk.proske@boku.ac.at

DOI 10.1007/978-3-540-77618-5

Springer Heidelberg Dordrecht London New York

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is

concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,

reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication

or parts thereof is permitted only under the provisions of the German Copyright Law of September 9,

1965, in its current version, and permission for use must always be obtained from Springer. Violations

are liable to prosecution under the German Copyright Law.

The use of general descriptive names, registered names, trademarks, etc. in this publication does not

imply, even in the absence of a specific statement, that such names are exempt from the relevant protective

laws and regulations and therefore free for general use.

Preface

Stone arch bridges are special technical products in many aspects. Two of

the most important aspects are their very long time of usage and their land-

scape changing capability. First, for more than two millenniums, stone

arch bridges have been part of the human infrastructure system, and some

of them are still in use. Most of the stone arch bridges now in use are older

than the first century. The only type of structures reaching the same dura-

tion of usage are tombs and other religious structures. However, in contrast

to those, arch bridges are much more exposed to changes in usage condi-

tions. There exist Roman bridges that were crossed not only by Roman

legions but also by tanks in World War II. When most stone arch bridges

were constructed, motorized individual car traffic was yet unknown. This

load now has to be borne by these historical bridges. We should probably

much more esteem the farsightedness and endeavour of our ancestors,

which we often count on nowadays without perception.

Or perhaps we do notice as some common attitudes indicate, don’t we?

In many children’s books, landscapes often include stone arch bridges.

And if people are asked whether arch bridges are disturbing or accepted, in

most cases people consider arch bridges as part of our man-made land-

scape and not necessarily as human artefact. Painters such as Paul Cézanne

have included arch bridges in their landscape paintings as early as the

19th century, which refutes the theory that arch bridges are now just

accepted because they have been part of the landscape for centuries.

Stone arch bridges are considered beautiful because they apply some

simple rules of aesthetics. First of all, they use building material from the

vicinity and therefore are embedded in the landscape. Furthermore, the

genius idea to arrange stones geometrically in such a way that the mechani-

cal properties of stones are used in a nearly perfect way gives the impres-

sion of harmony, whereas beam bridges made of reinforced or prestressed

concrete are often felt as strange.

Besides beauty, the bridges show in a very clear way one of the biggest

conflicts of our human civilisation. In the untiring trial to rationally de-

scribe all elements of our world, we have seen the limits of this concept in

the last decades. Even though arch bridges have been built and used for

more than two millenniums, we still face problems in numerically describ-

ing their behaviour. Only in the last decades have appropriate tools been

VI Preface

provided. Such tools are presented in this book. However, the book em-

beds these procedures in an even wider concept. Not only are computation

strategies and strengthening techniques for arch bridges given, but adapta-

tions of today’s loads to preserve the bridges are also presented.

However, strengthening of arch bridges is often not required: The major

cause of the destruction of arch bridges is the insufficient width of the

roadway, which means not the safety but the usability has limited the life-

time of the bridge. Perhaps we could live with this limitation and give re-

spect to the arch bridges. They still provide us with the lowest maintenance

costs of all bridge types.

Expression of Thanks

marily written during a visit of the first author at the TU Delft in 2005.

Therefore, first of all we thank the TU Delft for the financial support. Ad-

ditional support was given to translate the book.

Furthermore, many persons have contributed by proof-reading and giv-

ing suggestions. Therefore, the authors thank Prof. Konrad Bergmeister,

Prof. Han Vrijling, Prof. Jürgen Stritzke, and Prof. Udo Peil. Additionally,

Mrs. Angela Heller did the proof-reading of the German version. We thank

her for the intensive work. Last, but not least, we thank Springer and their

assistants for the opportunity to publish the book in English, and for their

strong support.

Contents

1 Introduction........................................................................................... 1

1.1 General introduction ........................................................................ 1

1.2 Advantages and disadvantages of arch bridges ............................. 12

1.3 Structure of the book ..................................................................... 14

1.4 Terms ............................................................................................. 20

1.5 Classification of static bridge types ............................................... 29

1.6 Types of arch geometry ................................................................. 33

1.7 History of stone arch bridges......................................................... 36

1.8 Arch bridges from alternative material.......................................... 47

1.8.1 Steel arch bridges ................................................................. 47

1.8.2 Wooden arch bridges ........................................................... 48

1.8.3 Concrete arch bridges........................................................... 49

1.9 Number of arch bridges ................................................................. 50

References ............................................................................................ 56

2 Loads.................................................................................................... 67

2.1 Introduction ................................................................................... 67

2.2 Road traffic loads........................................................................... 67

2.3 Railroad traffic load....................................................................... 82

2.4 Initial drive forces.......................................................................... 86

2.5 Breaking forces.............................................................................. 87

2.6 Wind loading ................................................................................. 88

2.7 Impact forces ................................................................................. 89

2.8 Settlements..................................................................................... 89

2.9 Temperature loading...................................................................... 89

2.10 Snow loading ............................................................................... 92

2.11 Dead load..................................................................................... 92

References ............................................................................................ 94

3.1 Introduction ................................................................................... 99

3.2 Empirical rules............................................................................. 100

3.2.1 Historical rules................................................................... 100

3.2.2 Modern rules...................................................................... 119

X Contents

3.3.1 Single beam models........................................................... 125

3.3.2 Compound beam models ................................................... 133

3.4 Finite element method (FEM) ..................................................... 136

3.5 Discrete element method (DEM)................................................. 140

3.6 Comparison of testing and modelling.......................................... 141

3.6.1 Load tests on arches........................................................... 141

3.6.2 Comparison results ............................................................ 144

3.7 Transverse direction (effective width)......................................... 148

References .......................................................................................... 152

4.1 Introduction ................................................................................. 165

4.2 Masonry elements........................................................................ 166

4.2.1 Masonry stones.................................................................. 166

4.2.2 Mortar................................................................................ 172

4.3 Maximum centric masonry compression strength ....................... 176

4.3.1 Model according to DIN 1053-100 ................................... 177

4.3.2 Model according to DIN 1053........................................... 178

4.3.3 Empirical exponential models ........................................... 179

4.3.4 Model according to Hilsdorf ............................................. 179

4.3.5 Model according to Mann ................................................. 180

4.3.6 Model according to Berndt................................................ 181

4.3.7 Model according to Sabha ................................................. 183

4.3.8 Model according to Ohler.................................................. 183

4.3.9 Model according to Stiglat ................................................ 184

4.3.10 Model according to Francis, Horman and Jerrems.......... 184

4.3.11 Model according to Khoo and Hendry ............................ 184

4.3.12 Model according to Schnackers....................................... 185

4.3.13 Model according to Ebner ............................................... 185

4.3.14 Further masonry compression models............................. 185

4.4 Stress-strain relationship.............................................................. 186

4.5 Moment-Axial force diagrams..................................................... 187

4.6 Additional-leaf masonry .............................................................. 188

4.6.1 Introduction ....................................................................... 188

4.6.2 Model according to Warnecke........................................... 188

4.6.3 Model according to Egermann .......................................... 188

4.7 Shear strength .............................................................................. 190

4.8 Proof equations ............................................................................ 191

References .......................................................................................... 192

Contents XI

5.1 Introduction ................................................................................. 199

5.2 Destructive tests........................................................................... 202

5.3 Semi-destructive test methods ..................................................... 205

5.4 Non-destructive test methods ...................................................... 205

5.4.1 Ultrasound ......................................................................... 205

5.4.2 Impact-echo ....................................................................... 206

5.4.3 Radar ................................................................................. 207

5.4.4 Tomography ...................................................................... 208

5.4.5 Thermography ................................................................... 208

5.4.6 Electrical conductivity....................................................... 208

5.4.7 Experimental tests on bridges on site ................................ 208

5.4.8 Photogrammetry and lasercanning .................................... 209

References .......................................................................................... 210

6.1 Introduction ................................................................................. 217

6.2 Damages on historical arch bridges ............................................. 218

6.2.1 Overview ........................................................................... 218

6.2.2 Recent collapses of historical arch bridges........................ 221

6.2.3 Weathering of the mortar................................................... 222

6.2.4 Spalling and contour scaling ............................................. 223

6.2.5 Salt attack .......................................................................... 224

6.2.6 Chemical weathering ......................................................... 226

6.2.7 Biological weathering........................................................ 226

6.2.8 Mechanical and physical weathering................................. 226

6.2.9 Deformations ..................................................................... 227

6.2.10 Cracks .............................................................................. 229

6.3 Repair and strengthening ............................................................. 235

6.3.1 Introduction ....................................................................... 235

6.3.2 Strengthening techniques................................................... 241

6.3.3 Examples ........................................................................... 254

6.4 Arch bridges of the second generation ........................................ 256

References .......................................................................................... 257

7.1 Definition of safety and safety concepts...................................... 265

7.2 Probabilistic safety concept ......................................................... 269

7.2.1 Introduction ....................................................................... 269

7.2.2 First-order reliability method (FORM).............................. 270

7.2.3 Second-order reliability method (SORM) ......................... 274

7.2.4 Hypersphere division method............................................ 281

7.2.5 Response Surface Method ................................................. 281

XII Contents

7.2.7 Combination of safety indexes .......................................... 288

7.2.8 Limitation of the presented methods ................................. 293

7.2.9 Commercial programs ....................................................... 295

7.2.10 Goal values of safety indexes .......................................... 297

7.3 Semi-probabilistic safety concept................................................ 301

7.3.1 Introduction ....................................................................... 301

7.3.2 Partial safety factors .......................................................... 303

7.3.3 Characteristic values.......................................................... 310

References .......................................................................................... 323

8 Examples............................................................................................ 333

8.1 Introduction ................................................................................. 333

8.2 Examples in literature.................................................................. 333

8.3 Own examples ............................................................................. 336

8.3.1 Bridge 1 ............................................................................. 336

8.3.2 Bridge 2 ............................................................................. 341

8.3.3 Bridge 3 ............................................................................. 345

8.3.4 Bridge 4 ............................................................................. 346

8.3.5 Bridge 5 ............................................................................. 348

8.3.6 Bridge 6 ............................................................................. 349

8.3.7 Summary ........................................................................... 350

8.4 Further examples ......................................................................... 351

8.4.1 Historical stone beam bridges ........................................... 351

8.4.2 Anchoring chamber of the Blue Wonder Bridge............... 352

8.5 Conclusion ................................................................................... 354

References .......................................................................................... 357

1 Introduction

“The first bridges men built were in wood, which were suited to their re-

quirements at the time. But then they began to think about the immortality

of their names. And because their richness gave them heart and made bet-

ter things available to them, they began to build bridges in stone, which

lasted longer, cost more, and brought glory to those that built them.”

technology, which is seen as a real characteristic of human culture, already

shows the following fact by its definition: under technology, one under-

stands the method and capacity to use the natural phenomena in a practical

way.

This statement also applies to the technical product of bridges. Nature is

easily able to create bridges without human involvement: bridges, as

physical features, already existed for millions of years, created from geo-

logical formations by wind and water or as fallen trees that cross a creek.

One only has to travel to the Arches National Park in the United States to

see examples.

However, at the moment in which nothing except the forces of nature

create bridges without the forces of intellect, the situation changes funda-

mentally. An artificial phenomenon then originates. The artificial distin-

guishes itself from the natural by the agreement of intention. Thus, a line,

for example, becomes an artificial phenomenon, when the lines are shaped

into a symbol. This symbol, however, requires an agreement in advance.

Based on this hypothesis, a bridge is considered to be an artificial phe-

nomenon since it is designed to span something and to ease progress.

Next to that, it seems as if the concept of art is related to the notion of

“artificial”. But the notion of art is actually more strongly linked to the no-

tion of technology. Originally, the word “art” describes a high degree of

D. Proske, P. van Gelder, Safety of Historical Stone Arch Bridges, DOI 10.1007/978-3-540-77618-5_1,

© Springer-Verlag Berlin Heidelberg 2009

2 1 Introduction

skill with which a human accomplishes a task. Later on, the term is ex-

panded to the product itself.

An example of a product that deserves the notion of artwork is the world

atlas “Atlas Maior” (later edition 2005) that appeared between 1662 and

1665. The atlas showed the geography of the known world in a quality that

had never been achieved until that time. The publisher from Amsterdam Joan

Blaue writes in the preface of the enclosed maps: “[With them ...] we tread

inaccessible mountains and traverse oceans and rivers without risk” (FAZ

2005).

The human insistence to explore, which drives one to expand his own

living space, eventually cannot be satisfied by maps or by pure conceptual,

context only. But these maps can be a movement towards it. The spatial

movement of humans and objects does have a substantial meaning, not

only in science, but also in the present day-to-day world.

Since the beginning of humanity, humans could only cover large spa-

tial distances in a very long time frame. The individual could not separate

himself too far away from his place of origin. Yet the settling of humans

for 10,000 years after the last Ice Age on the necessity of agricultural

grounds once again contradicts the human wish for mobility. However,

settlement led to an unexpected effect: it paid off by making “inaccessible

mountains” crossable and “oceans and rivers” traversable without risk,

since this has to be done on a regular basis.

The broadening of the skill of the human movement apparatus with re-

gard to the development of speed, with regard to the development of dy-

namic mingling, and with regard to the reach has probably lead to the de-

velopment of improved movement and transport systems, respectively,

with the start of settlement.

The invention of the wheel, as well as The taming of horses, is often re-

garded as one of the greatest inventions of humankind, because there exists

no example in nature of a wheel that rotates around its own axis. The

wheel is the basis for the construction of vehicles that allow the transport

of goods and people in a very economic way. The basis for the contribu-

tion of wheels in transport is, however, a proper condition of the surface of

the roads. One of the requirements is a certain wheel straightness and hori-

zontalness of the roll face. Such wheel straightness also relieves the human

and animal movement apparatus. This is meaningful, because pulling ani-

mals like horses served as the power for vehicles for almost 5,000–6,000

years after the invention of wheel.

The free and efficient movement of vehicles, people, and animals is, for

example, not given for the movement through rivers. Also, very steep roads

1.1 General Introduction 3

quickly result in the exhaustion of both humans and animals and are un-

suitable for wheeled vehicles. Therefore, the wish to ease the movement of

people and goods arose probably very early. The best way to ease move-

ment is shortening and horizontally or vertically detouring to avoid unsuit-

able stretches of the road. Bridges follow this idea. They are structures that

serve to cross obstructions. They not only can be used for various means of

transportation (road and railway bridges) for people (pedestrian bridges)

and animals, but also for passing over water.

When one compares bridges with the above-mentioned maps, they are

both invitation and tool for movement in a double sense: not only do you

show the design of the roads, just like on the map, but you also offer a

physical entrance. The first examples of primitive bridges are the stone

bridges of Dartmoor (Brown 1994), and the stone beam bridges in Gizeh

(2,500 B.C.) in China (500 B.C.) (Heinrich 1983).

The physical achievements brought about by the creation of bridges are

substantial. Even in present times, with the elaborate technical resources

available, many people sense an inner feeling when they marvel at the

beauty and size of the extraordinary historic bridge structures. This is es-

pecially true for the mighty arched bridges of the Romans—for example

the Pont du Gard in France or the Bridge of Alcántara in Spain. Many of

these over 80-generations old structures, like the Ponte Milvio, were

crossed by Roman legions as well as by German and American armoured

vehicles in World War II (Heinrich 1983, Straub 1992). These military op-

erations basically always have the objective of demolition or destruction.

In contrast to this, a bridge is a construction – a synthesis. In this pas-

sage, it is noted that humankind is also a synthesis. The translation of the

concept “synthesis” in Latin is composition, also additio (von Hänsel-

Hohenhausen 2005). The written work at hand is the composition of an

analysis. An analysis again is a derivation, since the Latin word for it is

reductio (von Hänsel-Hohenhausen 2005). The objective of this analysis,

however, is the progress of the actual structure – the progress of the syn-

thesis.

Progress can reach far over the everyday present. It will and must en-

compass future generations. Our ancestors, for example, encouraged us as

they erected bridges that are useful even today.

The preservation and acknowledgment of the skills needed to erect

historic bridges is, in the view of the writers, a duty we have to the up

coming generations. One can best carry out this duty when one defines a

use for the historic bridge structures and extends their utilization time.

Just as work is an inextricable criterion for the merit of a human being,

the utilization of a bridge is an inextricable criterion for the justification

of its existence.

4 1 Introduction

These structures, however, are only utilized when the advantages from

its utilization are greater than the possible disadvantages. An elementary

disadvantage of a historic structure can be the poor capacity for present

day loads. Complying with modern safety codes is not negotiable for any

product, even for historic structures. The mentioned safety requirements of

technically created structures are transposed through safety concepts. In

recent years, a new safety concept for structures has been introduced both

in European context and also on a national level in Germany.

The successful integration of historic arch bridges of natural stone into

these safety concepts is the foundation for a reutilization of these types of

bridges. This book takes up that task.

Figures 1-1 to 1-15 not only provide an impression of the great diversity

of masonry arch bridges, but also arch bridges of other materials. How-

ever, many further fine examples are known. The reader can consult the

homepages by Bill Harvey (2006) or Janberg (2008). Another interesting

example is the Minzhu Bridge in China (NN 2005), where three half

arches meet in the arch crown. Some better modern examples are the com-

pletion of the stone arch bridge Pont Trencat in Spain by a steel arch with

closed spandrel walls, a bridge in the inner port of Duisburg with a flexible

lane that is lifted up in case of ship traffic and thus can perhaps count as an

arch bridge (Bühler 2004), the Gateshead Millennium Steel arch bridge in

Newcastle by Chris Wilkinson with a curved lane that rotates along its

alongside axis in case of ship traffic, the Puente La Barqueta (Langer’s

Beam), the steel arch bridge with suspension above Ebro in Logrono, the

Leonardo Bridge in Norway, or the Juscelino Kubitschek Bridge in Brazil

(Goldberg 2006). Besides the success of steel and concrete arch bridges, in

recent years, some stone arch bridges have again been erected in Great

Britain and Portugal.

As the above-mentioned examples demonstrate, the types of arch

bridges change and live on by continuous variation of the original idea.

The presently applied mathematical optimization for defining optimal

bridge variants, leading to uniform standard solutions in the end, will and

has already partially failed. One such optimization requires a multitude of

entry sizes, which are not yet known at the time of the optimization calcu-

lation. Next to this uncertainty, various optimal solutions are often possi-

ble. A magnificent example of this is the variety of living organisms on

Earth. One occasionally happens to find similar organic solutions, for ex-

ample for the limbs, but even so one frequently finds differences for the

same boundary conditions. The constant variation of the arch bridges is

one key element of their success.

1.1 General Introduction 5

Fig. 1-2. Sweden Bridge in Dresden, built in 1845 (view on the carriageway)

6 1 Introduction

Fig. 1-3. Ponte Vecchio and Ponte St. Trinità in Florence, Italy

1.1 General Introduction 7

8 1 Introduction

1.1 General Introduction 9

10 1 Introduction

1.1 General Introduction 11

12 1 Introduction

All biological solutions and technical systems have advantages and disad-

vantages. An advantage of arch bridges is that their beauty is certainly not

to be underestimated. The beauty of arch bridges is often traced back, not

only to the use of natural materials (stone look), but also to the high aestheti-

cal measure. Birkhoff (1933) has introduced such a concept. He defines this

aesthetical measure as follows,

O (1-1)

A=

C

in which the aesthetic measure A is between zero and one, the O stands for

the number of relations of order, and the C stands for the complexity. An

application of the measure can be found in Staudek (1999). An essential

assumption of this measure is the relationship between beauty and effec-

tiveness. According to Piecha (1999), humankind possesses assessment

1.2 Advantages and Disadvantages of Arch Bridges 13

ture. Objects are aesthetically surveyed after such an observation and as-

sessment mechanism. Birkhoff assumes that a high measure of aesthetic

satisfaction exists with a balanced relation between the observation effort

to identify orders within an object and the complexity of the object—for

example, the new elements. In case of arch bridges, the number of ordering

relations, as well as the complexity, is very limited, so arch bridges are

considered aesthetic in accordance with this consideration.

This fact fits very well with some observations. For example, in Switzer-

land, arch bridges actually act as tourist attractions. For the largest part of

the population, they are regarded as an element instead of an interference

with nature. This could admittedly also be because many historic arch

bridges have high seniority and with that a customary right. The above-

mentioned consideration is also strengthened, however, by the fact that

arch bridges are constructed on many historic sites and with that are then

already regarded as aesthetic. The French painter Paul Cézanne included

arch bridges in his paintings and considered the bridges as part of nature

(Becqué 1983)

But at this point it should be pointed out that the ultimate numerical de-

scription of beauty has not been achieved up to now. However, develop-

ments of the Birkhoff measures have taken place, for example, by Bense

(Ebeling and Schweitzer 2002, Klein 2008).

Further advantages of arch bridges, in addition to their indisputable

beauty, are summarized by Weber (1999):

• Limited deformations under traffic loads (some tens of millimetres in

case of railroad bridges)

• Usability and fatigue are irrelevant (the total strains are often in the cy-

clic pressure load region)

• Application of uninterrupted rails based on limited deformations (no rail

fissures required)

• A high failure safety and robustness (insensitive to unplanned impacts)

• A high damage tolerance (recently, the notion of fitness is also used for

that: a system has high fitness when it stays functional despite a large

number of occurring faults)

• Early indication of malfunctioning

• A long lifetime and period of utilization

• Arch bridges, like all deck bridges, guarantee an undisturbed view for

the travellers

• The construction materials can be disposed of and re-used as environ-

mentally compatible material, respectively

• Excellent insertion into the landscape

14 1 Introduction

• Considerable reduction of the loading capacity by large support dis-

placements (this assumption is however valid for all bridges)

• The clearance diagram under the bridge is not constant

• Complex renaturation

Obviously, such a summary of pros and cons is always subjective. It ap-

pears, however, as if the advantages prevail over the disadvantages. That

would support the preservation of these structures, and with that, the reali-

zation of safety assessments with the aim of conservation.

Petryna (2004) has displayed a very good description of the basic elements

of damage-oriented safety analyses of structures in his paper (Fig. 1-16).

The description shows the following five basic elements: load models, ma-

terial models, damage models, carrying models, and the formulation of

verification equations. These elements can be found in this book, as well.

Figure 1-17, according to Mori and Nonaka (2001) and Mori and Kato

(2003), is another way of describing the time-dependant probability of

failure of a structure as a safety measure substitute without, however,

considering the load and carrying models.

The relatively abstract description by Petryna (2004) can be transformed

into a four-staged assessment scheme according to Diamantidis (Fig. 1-18)

and ICOMOS (2001) (Fig. 1-19). Then, the single points of the process as

well as the personal allocation are mentioned in this scheme. There is

also a less detailed approach for that purpose by Czechowski (2001)

(Fig. 1-20).

Figure 1-21 classifies the safety assessment of bridge maintenance pro-

cedures. However, the maintenance procedures for historic monuments

under protection include some further restraints (ICOMOS 2001, Vockrodt

2005, Vockrodt et al. 2003, Yeomans 2006, Žnidarič and Moses 1997,

Rücker et al. 2006, Jensen et al. 2008, and SIA 269 2007).

That the reliability theory-based analysis can ultimately only be a part of

an over-organized and discipline-crossing inspection and maintenance

strategy is shown in Fig. 1-22. It clarifies the various historic develop-

ments of risk-based inspection concepts for different industrial areas.

Whereas in some areas, (for example risk-based inspection and mainte-

nance) is standard practice in the monitoring of offshore platforms, this

approach has not yet been implemented into other areas such as the con-

struction industry. This is for various reasons—i.e., different owners.

1.3 Structure of the Book 15

structures such as oil platforms normally belong to commercial firms. For

the state, meeting safety responsibilities is of utmost importance, whereas

commercial firms are also constrained by market-based forces. Therefore,

safety responsibilities are only a subtask. This is subsequently shown in

cost-efficiency considerations for safety precautions.

Dealing with protective measures, however, is and stays a political deci-

sion. The practically active engineer cannot argue with such political deci-

sions in his daily job. Structuring as in Fig. 1-18 would certainly be ideal

for this situation. Despite this, the systematic by Petryna (2004) is chosen

in this chapter, since it is only slightly bounded by organizational limita-

tions. Before the single emphases are discussed, however, this chapter first

gives some general information regarding arch bridges.

according to Petryna (2004)

16 1 Introduction

design. Just as for historic structures, extraordinary large distinctions can also ex-

ist here, insofar as the description of the empirical calculation principles for his-

toric arch bridges is necessary to estimate this value.

d Structurally converted load capacity. Not only structural modifications, but also

deficiencies during construction are considered here.

e Time and damage-dependent development of the loading capacity.

f Restoration of the loading capacity by a maintenance measure. One such meas-

ure can result in a partial, a complete, or an improved loading capacity. Some

cases are known, however, in which such maintenance measures resulted in an ac-

celeration of the damage development—i.e., by the use of the wrong mortar—

which leads to damage to the natural stones because of a greater stiffness.

g Effects of dead load.

h Effects of upgrading loading. A simplified assumption is made here that in

time, for example, the parapets are converted, backfills are exchanged, and spare

vaults are filled up.

i Continuously changing effects, such as wind, These effects are normally not

decisive for the massive arch bridges. However, traffic loading for continuously

used bridges can also be reckoned among this type of impact. These can then defi-

nitely become dominant.

j Impulse type effects, such as bumps and hits. Normally, exceptional effects are

dealt with here.

Fig. 1-17. Representation of the time-dependent probability of failure for safety

assessment based on Mori and Nonaka (2001) and Mori and Kato (2003)

1.3 Structure of the Book 17

(2003), Diamantidis (2001) and Schneider (1996)

18 1 Introduction

(2001)

1.3 Structure of the Book 19

Fig. 1-21. Flow chart of bridge inspection, evaluation, and strengthening accord-

ing to REHABCON (2000)

20 1 Introduction

ferent industries according to Goyet (2001)

1.4 Terms

The German term for arch “bogen” can be traced back to the old high

German “bogo.” It can also be found in the Dutch “boog” and the English

“bow.” In the German dictionary by the Grimm brothers from the 1800s, it

is called, following Kurrer (2002): “Arch now is the curved, bended,

twisted.” The root of the German term for arch (bogen) lies in the verb

bending. “According to Kurrer (2002), an arch is a concave curved support

framework from double flexural rigid building materials, in a structural

sense.”

Pauser (2003) writes on arch bridges: “Arches derive their high struc-

tural efficiency from the utilization of the compression cross section.”

Pauser (2002) defines an arch as “a curved compression member with a

transverse load.”

1.4 Terms 21

Weber (1999) defines: “An arch appears when a rod shaped (line

shaped) support framework, whose system line is located underneath its

tangents, is also concave shaped. The support framework experiences two

translatory movement restraints in each support in its main curvature

plane. Its construction materials are pull, push and pressure capable.”

In contrast to that word exists the term “vault.” The origin of the term

“vault” (German: Gewölbe) probably lies in the Roman term “camera.”

This term is applied to curved ceilings and eventually not only to the ceil-

ings themselves but also to the space below the ceiling: “... camera became

the broad term for the entire room that is covered by the ceiling.” This is

how it is called in the Grimm dictionary (following Kurrer 2002). Later on,

the term vault was ascribed back to the support structure again. Already in

1735, the following term definition can be found: “a ceiling shaped from

an arch of stones.” In 1857, the term was extended to other materials as

well (Kurrer 2002).

The definition that domes carry out their support function only by com-

pression-capable construction materials with negligible tensile resistance

was occasionally disputed. But as the following examples prove, the defi-

nitions of the term “vault” display a great divergence.

Haser and Kaschner (1994) define: “With vault bridges, whelmed sup-

port structures are considered with an in front view curved support axis,

which geometrically demonstrate a surface shape based on their limited

cross-sectional height compared to their large cross-sectional width and

with that distinguish themselves from the rod shaped considered arch

bridges.”

Lueger (from Weber 1999) defines: “A vault is a stone ceiling assem-

bled from wedge-shaped stones that as a result impends freely and trans-

fers its operating loads and own weight to walls and columns.”

Mörsch (from Weber 1999) defines: “The vault bridges can be regarded

as arched girders from a static point of view, because they exert a horizon-

tal push as a result of vertical loads.”

Dimitrov (from Weber 1999) defines: “Vaults are arched girders, whose

statue is based on the pressure line.”

Kurrer (from Weber 1999) defines: “A support structure is a vault when

the support function, required as safeguard for traversing a space, is only

realized by compression proof construction materials with a negligible ten-

sile resistance.”

Weber (1999) defines: “A vault bridge appears as a support structure to

cross over roads and obstacles. This support structure is characterized by a

curved system surface with either only parabolic or parabolic and elliptic

points out of the sides plus a clearance of at least 2.0 m. Its material is

compression capable with a negligible small tensile resistance.”

22 1 Introduction

hardly noticed. There are various definitions for culverts as special types

of vaults and arch bridges, respectively. According to Mörsch (1999), cul-

verts can be distinguished by small spans (<8 m), by a high earth fill above

the key, and by a high rise–span ratio (f/l > 1/3). Other publications state a

span of 2 m (Orbán 2004) or 3 m (Bién and Kamiński 2004). Mörsch

(1999) furthermore discerns river and valley bridges from vaults. River

bridges thus distinguish themselves by flattened arches and greater spans,

and valley bridges by high columns and semi circular-shaped arches.

The terms for the single components of a vault bridge are displayed in

Figs. 1-23 and 1-24. English terms and definitions of arch bridge elements

can be found in Griefe (2006). A radial array of stones is regarded as a real

vault, whereas a false vault exists of cantilevers.

The integration of vaults and arch bridges, in the classification of

bridges respectively, follows in the next section. The various kinds of vault

bridges are displayed in Fig. 1-25.

Fig. 1-23. Elements of stone arch (vault) bridges according to Huges and Blackler

(1997)

1.4 Terms 23

Keystone/Crown

Fig. 1-24. Terms of stone arch (vault) bridges according to Koch (1998)

Further terms for special types of bridges are only touched upon hereaf-

ter, because they are reserved for education on window arches. A stilted

arch has an elongation between curvature and springing. The springings lie

at different heights in case of rising or single-hip arches. Further notations

for arch bridges are round arches, flat bow, rising, segmental arches, ellip-

tical arches, basket arches, shoulder, collar plunge, panel, pointed, lancet

arches, clover leaf or trefoil arches, fan, jagged, keel, saddle-backed cop-

ing, flame arches, curtain arches, tudor arches, horseshoe arches, and plunge

arches. The arch shapes for vaults bridges will still be treated later on.

24 1 Introduction

Fig. 1-25. Types of arches according to Koch (1998)

1.4 Terms 25

overpassing: “Aquaduct.” This term stems from the Latin term aquae ductus

= water conduit. This refers to antique Roman constructions that, as arch

bridge, often transport an open or a closed water channel to a settlement.

The fillings of a vault bridge lying between the vault and the roadway

are denoted as backfilling. Backfilling can be developed differently, both

constructively and in the choice of materials. The backfilling can thus con-

sist of

• unbound imported fill (loose material)

• reinforced soil by means of injection

• concrete or walled support

• hollows for weight saving, alongside arranged fittings (front and between

alongside the walls) (Haser and Kaschner 1994)

With many bridges, one has attempted to achieve weight saving by hol-

lows and openings in the backfilling. Thus, spare vaults are incorporated in

numerous bridges. These spare vaults can run both in longitudinal and

transverse directions. While an example of spare vaults in longitudinal di-

rection is the Orleans Bridge by Perronet (Fig. 1-26), an example of spare

vaults in transverse direction is the Verde Bridge in Italy (Fig. 1-27). The

spare vaults can be arranged differently, often parallel, sometimes on top

of each other, or even irregularly. In many cases, the spare vaults are cov-

ered by spandrel walls. As a rule, fully closed spandrel walls are, however,

not incorporated for spans greater than 70 m to avoid unwanted stiffening

by the spandrel walls. Melbourne and Tao (1995, 1998, 2004) are men-

tioned as further literature for bridges with open spandrel walls.

Towards the end of the 19th century, however, spare vaults were again

increasingly disguised by spandrel or front walls. In Italy, one presumes

that all bridges with a span over 20 m contain spare vaults in an arbitrary

form. Towards the end of World War II, spare vaults were only very sel-

dom incorporated. As a rule, the bridges were entirely backfilled, because

the labour costs for the construction of spare vaults were greater than the

material costs (Brencich and Colla 2002).

One of the first bridges with longitudinal spare vaults was the Westmin-

ster Bridge in London. In 1748, it was decided, after settling occurred at

Pillar 4, to construct adjacent arches with smaller vaults and with the ap-

plication of spare vaults with a lower deadweight to decrease the loads on

the foundations (Brencich and Colla 2002).

26 1 Introduction

Fig. 1-26. Longitudinal spare vaults after Brencich and Colla (2002)

Fig. 1-27. Transverse spare vaults in the Verde Viaduct in Italy, constructed

1883–1889, span 18.5 m, after Brencich and Colla (2002)

Although spare vaults with very flat arches, with elliptical arches, and

even with pointed arches can occasionally be found, spare vaults are usu-

ally constructed as half-circular arches to limit the horizontal loads on the

walls of the spare vault, especially with the front walls as the outer wall. In

the case of very flat arches (segmental arches) in the spare vault, metal

chains are used halfway to transfer the horizontal loads. The layers of the

spare vault are chosen in such a way that the loads on the walls of the

spare vault generate a compression arch line that follows the main arch

(Schaechterle 1937, 1942).

The application of multi ring arches in the main arch at the connection

to the spare vaults occasionally leads to separation of the arches. This con-

struction solution was frequently applied at the end of the 19th and the

1.4 Terms 27

of arches can be found at the Badstrassen Bridge in Berlin. This bridge

was constructed between 1861 and 1864. It involves a skewed multi ring

bridge across six spans. In two spans, the skewness is achieved through the

shifted string of straight sectors of arches. These arch sectors are delimited

in contact sector to the nearest row of arches by the front walls. Addition-

ally, a front wall is located in the middle of such a sector of arches, as well.

The front walls end at the springing sector. The connection between the

front walls of the single arrays is achieved through spare vaults across the

springing. These spare vaults are also filled with scrap rock and mortar. A

minimal weight for simultaneous regular support of the main arch is ac-

complished by the combination of stiffening walls and filled spaces. Next

to the application of spare vaults in the backfilling, they are also regularly

used in the pillars (Brencich and Colla 2002).

The possibility of multi ring arches must also be checked when no multi

ring arches are visible from the outside of the arches. Especially in the

second half of the 19th century through the thirties of the 20th century, the

multi ring arches were often disguised by natural stones in the edge region.

Thereby, the impression of a thicker arch is created. Actually, significant

distinctions can appear between the apparent thickness of the arches and

the actual thickness. The Cornigliano Bridge (Italy) from the year 1932 is

mentioned here as an example. Additionally, the thickness of the arches in

the inner sector can often indicate significant variances, since stones are

arranged with varying accuracy and rise up to various depths in the back-

filling (Brencich and Colla 2002).

Also, the springing construction outside and inside often points out dis-

tinctions, as shown by the example of the Verde Bridge in Italy in Fig. 1-28.

Further examples of the distinction between visible and structural compo-

sitions are shown in Figs. 1-29 and 1-30.

After this rough introduction to some of the arch bridge elements, this

bridge type shall be integrated into a general bridge system.

28 1 Introduction

Fig. 1-28. Construction of abutments in the Verde Viaduct. The left image shows

the visible springing and the right image shows the actual inner construction as

multiple shelled stonework arches with overhanging springing stones (Brencich

and Colla 2002).

Fig. 1-29. Construction from the springings at semi elliptical arches (from Brencich

and Colla 2002). Visible springings must not coincide with the static springing

1.5 Classification of Static Bridge Types 29

Fig. 1-30. Examples of a pillar construction after Brencich and Colla (2002)

Schlaich (2003) has introduced a typology of bridges as shown in Fig. 1-31.

Historically, the statical systems were determined by the mechanical prop-

erties of the building material, usually taken from the vicinity. However,

with the development of new building materials, especially steel and com-

posite materials such as reinforced concrete, the possibilities for bridge de-

sign increased and traditional limits were exceeded. The longest arch

bridge worldwide is the Lupu Bridge in Shanghai, China, with a span of

550 m (Chen 2008). Dubai plans to build the Bur Dubai-Deira arch bridge

with a span of 667 m (Chen 2008). In China, arch bridges with an incredi-

ble span of 1,000 m are planned (Čandrlić et al. 2004, Martínez 2004). Such

spans can only be achieved with advanced construction materials.

Whereas in historical times location and shape of the bridges were

strongly limited, nowadays architects and engineers experience more

freedom in the choice of bridge type by choosing from the tool box of

materials.

Historical building materials were axial tensile force-capable ropes, ax-

ial compression-capable masonry, and bending and axial force-capable

wood. However, wood geometries were limited by biological limitations,

for example the size of trees. Using these materials, suspension bridges

with ropes, arch bridges with masonry, and beam bridges with wood were

constructed for probably more than two millenniums.

30 1 Introduction

An arch bridge and a suspension bridge in these concepts are the assem-

bly of several single structural elements in such a way that the experi-

enced force types comply optimally with the capable force types. If such

an arch structure is designed, the location of the roadway can either be up

on the arch or suspended from the arch. The latter is not found for stone

arch bridges but for steel arch bridges. Furthermore, the so-called Langer’s

Beam is distinguished by not transfering horizontal loads to the foundation

but keeping the horizontal forces inside the roadway by tensile structural

elements.

However, in the context here, a more appropriate classification of arch

bridges is shown in Fig. 1-32, originating from Bién and Kamiński (2004).

The classification considers the building material, the arch geometry, the

arch thickness, the number of spans, and the type of the front or spandrel

wall.

Bridges can cross the obstacle either in rectangular form or in another

angle. Bridges crossing differently than in rectangular form are called

skewed bridges. Skewed masonry arch bridges are classified according to

the joints pattern (Figs. 1-33 and 1-34).

A detailed discussion of skewed, multi ring arch and multi span bridges

is not part of this book. For skewed bridges, please refer to Chandler and

Chandler (1995), Choo and Gong (1995), Melbourne (1998), and Hodgson

(1996); for multi ring arch bridges, please refer to Gilbert and Melbourne

(1995), Gilbert (1998), and Drei and Fontana (2001). For multi span

bridges, detailed information can be found in Molins and Roca (1998) and

Fanning et al. (2003).

Figure 1-31 shows that the arch shape is an important property for the

classification of stone arch bridges. Therefore, this property will be dis-

cussed in detail in the next section.

1.5 Classification of Static Bridge Types 31

32 1 Introduction

Fig. 1-32. Typology of stone arch bridges according to Bién and Kamiński (2004)

1.6 Types of Arch Geometry 33

springing method Method (Joints cut the

middle line of the arch

rectangular)

Fig. 1-33. Typology of joint patterns in skewed masonry arch bridges according to

Melbourne (1998)

In the section on terms, it was indicated that arches are curved and possess

a curvature. The selection of the curvature and the geometry of the arch is

34 1 Introduction

based on a maximum match between the line of thrust and the arch geome-

try. The line of thrust depends heavily on the type of loading. Therefore,

for example, a circular arch is the optimal choice for a constant radial load

(Fig. 1-35 left), a parabolic arch results from a constant vertical uniform

load (Fig. 1-35 middle), and a catenary arch is the optimum for constant

dead load of the arch (Fig. 1-35 right) (Petersen 1990).

Fig. 1-35. Optimal arch shape according to different loading patterns (Petersen

1990)

The term “line of thrust” describes the geometry under which the load is

only transferred by axial forces, here it is being transferred by compression

forces. It can also be compared to ropes, which transfer only axial tensile

forces. The affinity between ropes and arches also becomes visible by the

choice of the catenary arch geometry.

However, under realistic conditions, load changes and some further

conditions such as construction boundaries have to be considered when de-

signing the arch shape. Table 1-1 gives an overview about possible arch

shapes. The table mentions the circular arch, the parabolic arch, the ellipti-

cal arch, and the basket arch. The basket arch is a particular case of the cir-

cular arch since it is assembled from several circles with different radii. A

further arch shape is the cycloid. A cycloid is created as the trace of a point

inside a circle, when a circle is rolled on a certain line, usually a straight

line. It is therefore related to the circular arch. Formulas for the computa-

tion of certain geometries can be found in Petersen (1990). Weber (1999)

furthermore refers to Kammüller and Swida, mentioning a parable

fourth order for an earth backfilled arch bridge. And, last but not least,

Weber (1999) also mentions a sinus curve and a half wave as arch shape.

The identification of the exact arch curvature is often difficult, as shown

in Fig. 1-36. In Fig. 1-36, a historical reinforced concrete vault is shown.

The Santa Trinità bridge in Florence is chosen to illustrate the difficulties

in a second example. Ferroni assumed in 1808 that the arch geometry fol-

lowed a basket arch with six circular segments. Brizzi suggested two

parabolic arches in 1951, whereas Torricelli guessed a logarithmic

curve (Corradi 1998).

1.6 Types of Arch Geometry 35

Arch type Continuous arch Cross vault

Steep arch Low-pitched arch

Half-circular

arch or seg-

mental arch

Parabolic arch

Elliptic arch or

elliptic seg-

ment arch

Basket arch

historic epochs. For example, the Romans mainly used the half-circular

arch. However, the required rise-to-span ratio yielded a major bridge

height and consequently to long driveways with a significant slope. This

36 1 Introduction

caused certain problems, especially in cities. After the 16th century, more

and more basket arches, elliptical and centenary arches became popular,

since this shape of arches avoided the long driveways and slopes.

The application of arches and vaults for bridging space is probably several

thousand years old. Barrel vaults with a span of more than 1 m were al-

ready built about 5,000 years ago in Mesopotamic burial chambers (Kurrer

2002). Von Wölfel (1999) mentions the first known vault in the royal

grave of Ur about 4,000 B.C. Also, the Sumerians and the Old Egyptians

knew the vault (Heinrich 1983, Martínez 2004, von Wölfel 1999).

There are many different theories on how this type of structure was in-

vented. However, final proof of these theories is virtually impossible. One

theory claims that the overturning of false vaults yielded to the first arches.

Other theories consider the refinement of support stone elements or the

subdivision of stone beams into single elements as shown in Fig. 1-37

(Kurrer 2002, Heinrich 1983).

Van der Vlist et al. (1998) describes the development of arch bridges

from stone heaps over small creeks. Interestingly, Bühler (2004) mentions

the construction of natural bridges in the same way by Peruvian Indians.

Besides the mentioned first constructors of arch bridges, the old Greeks

also knew the fault and arch structure. However, they did not pay much at-

tention to it since they preferred strictly horizontal and vertical structural

elements. The Greek vaults were only applied late (after 350 B.C.) and

only used with small spans (less than 10 m) (Weber 1999). At that time, the

Greeks were not the only one to build vaults. Also, the Nabatäer on the

Arabian peninsula used vaults heavily for the covering of cisterns. Besides

that, the Nabatäer are well-known for the mountain city of Petra in Jordan.

A first major step in the development of experienced arch bridges was

during the time of the Etruscans. The Etruscans settled in the Northern

1.7 History of Stone Arch Bridges 37

middle of Italy before the time of the Roman Empire. The Etruscans have

been seen as the inventor of the wedge stone arch. In wedge stone arches,

every stone has a wedge-type shape, which allows a better shape of the

arch compared to ashlar-shaped stones. However, the Etruscans still did

not know mortar. And still, the placing of the stones in terms of adjustment

of joints towards the circle centre was mainly done with low quality.

A second step in the development of arch bridges was done during the

time of the Roman Empire. The Romans not only improved the quality of

the placement of the stones significantly, but also invented the mortar

and pentagonal-shaped stones to improve the link between the arch and the

spandrel walls. This permitted a radical improvement from wide vaults

built by the Etruscans to wide-spanned arch bridges with up to 36 m spans,

such as the bridge in Alcántara in Spain. The bridge about the Teverone

(nowadays Anio) close to Salario in Italy is one of the oldest stone arch

bridges of the Romans. The time of construction has been dated to 600

B.C. The bridge was destroyed about 1,000 years after constructions by the

East Goths (546 A.D.). The reconstruction was started in 569 A.D. The

bridge had a span of 22 m and a rise of 11 m.

The ratio of ½ for rise to span already shows a major problem of the

Roman arch bridges. Such high-rise bridges required steep ramps, which

were difficult to cross by carriages. A second problem of the Roman

bridges was wide piers. Here, the Romans developed countermeasures. Ei-

ther the piers were constructed with further openings to permit an extended

water flow in case of flooding or they simply positioned the superstructure

of the bridge very high above the valley. An example of the first technique

is the Fabricius Bridge in Rome. The bridge was probably constructed in

62 B.C. Interestingly, the Fabricius bridge is build on a complete circle:

the upper part of the circle forms the arch of the bridge and the lower part

of the circle forms an earth arch in the ground.

A further important Roman stone arch bridge is the bridge towards the

Engelsburg (Leonhardt 1982). This bridge was constructed around 137 A.D.

The bridge has long been considered one of the most beautiful bridges be-

cause when the water is smooth, the arch circle is closed by the mirror image

in the water. The Romans strived for such expression of harmony.

Due to the excellent stonecutter work and the good foundations, many

Roman arch bridges are still able to carry loads (von Wölfel 1999, Brown

1994, Zucker 1921, Jurecka 1979).

Gazzola (1963) published a catalogue of 293 known Roman arch bridge

structures (von Wölfel 1999, Leliavsky 1982). According to Weber (1999),

about 330 Roman arch bridges still exist. Most of these bridges are half-

circular arch bridges, and some of them are already segmental circular arch

bridges. In his book, O’Connor (1994) especially mentions the segmental

arch Pont St. Martin in Northern Italy, built around 25 B.C. and reaching a

38 1 Introduction

span of 35.6 m. Most Roman arch bridges were built between 50 B.C. and

150 A.D. (Gaal 2004).

The high quality of the Roman bridges was possible because of the

strong requirement of a sound Roman road system. Furthermore, the road

system was the basis for the existence of the Roman Empire itself. Accord-

ing to Fletcher and Snow (1976), the length of the road system reached a

length of 65,000 miles. The road system permitted a daily distance of

about 85 km. However, the Roman courier service “Cursus Publicus”

reached a daily distance of up to 335 km, with changing horses and couri-

ers which was an incredible value for that time.

Under Emperor Trajan (98–117 A.D.), the Italian provinces of the Roman

Empire had a road system of 16,000 km and about 8 million inhabitants.

Therefore, in the core regions of the Roman Empire, a ratio of 2 km road per

1,000 inhabitants was reached (Von Wölfel 1999). In comparison, Germany

currently has reached a value of 2.7 km per 1,000 inhabitants. Such high val-

ues could only be achieved by a strong economy. The Gross Domestic Prod-

uct of the Roman Empire was at least 10% above the average Gross Domes-

tic Product of the rest of the world at that time (Tables 1-2 and 1-3). The only

region with a comparably strong economy was China.

Not only road bridges were of major importance in the Roman Empire, but

viaducts—namely, another application of bridges—were also very important.

Several of the best known Roman bridges are viaducts, such as the Pont du

Gard (Garbrecht 1995). The viaduct of Segovia, probably constructed be-

tween 81 and 96 A.D., still reaches a distance of 818 m and the maximal

height of 28 m (Garbrecht 1995). If one believes a painting by Zeno Diemer

(Garbrecht 1995), then the intersection of Roman water supply pipelines such

as the Aqua Claudia/Aqua Anio Novus with the Aqua Marica/Tepula/Julia

was carried out at several levels and reminds one very much of modern

highway intersections. The Roman water supply reached values comparable

to modern water supply values (more than 140 litres per person per day).

Table 1-2. Per capita income in 1990 US dollars for different world regions (Streb

2003)

Average per capita income in 1990 US dollar in the year

Land/region 0 1000 1820 1998

Western Europe 450 400 1,232 17,921

USA, Canada 400 400 1,201 26,146

Japan 400 425 669 20,413

Latin America 400 400 665 6,795

Eastern Europe and USSR 400 400 667 4,354

Asia without Japan 450 450 575 2,936

Africa 444 440 418 1,368

World 444 435 667 5,709

1.7 History of Stone Arch Bridges 39

Table 1-3. Average economical growth in percent for different world regions

(Streb 2003)

Average economical growth in % in the years

Land/Region 0–1000 1001–1820 1821–1998

Western Europe –0.01 0.14 1.51

USA, Canada 0.00 0.13 1.75

Japan 0.01 0.06 1.93

Latin America 0.00 0.06 1.22

Eastern Europe and USSR 0.00 0.06 1.06

Asia without Japan 0.00 0.03 0.92

Africa 0.00 0.00 0.67

World 0.00 0.05 1.21

Besides the Romans, the Persians also constructed arch bridges over a

long period, presumably from 500 B.C. until 500 A.D. These arch bridges

were often elements of piled up dams. Some authors assume that Roman

constructions styles strongly influenced Persian bridge construction (von

Wölfel 1999).

Only shortly after the massive introduction of stone arch bridges by the

Romans arch bridges were also constructed in China. In contrast to the

Romans, which without exception built strong and massive arch bridges,

slender arch bridges were designed very early in China. The first bridge of

this type was the bridge in Luoyang, probably constructed 282 A.D. Since

then, many arch bridges were built in China, even arch bridges with sev-

eral spans. At least one of these bridges, the Jewel Belt bridge in Suhou

(built in 800 A.D.) is still in use. Furthermore, the Anji Bridge in China

should be mentioned here. Based on several authors, this bridge was the first

segmental bridge worldwide, designed and constructed in 605–607 A.D. The

bridge reaches a span of 37 m (Brown 1994, Yi-Sheng 1978, Jurecka 1979,

Graf 2005, Ding 1993 and Ding and Yongu 2001). Also, the Marco-Polo

bridge in China, built in 1194, is a worldwide-known historical stone arch

bridge in China (Brown 1994, Yi-Sheng 1978, Ding 1993 and Ding and

Yongu 2001).

After the decline of the Roman Empire in the middle of the first millen-

nium, the road system of the Romans degraded. This degradation also in-

cluded the bridges. The economy experienced a strong depression (Table

1-3). This is a very impressive example of the relation between bridge con-

struction and economical and social boundary condition (Heinrich 1983).

Only in the 11th and 12th centuries did a radical change reach Europe.

This change was accompanied not only by new technologies in agriculture

but also by the development of free bourgeoisie. The changes yielded to an

improvement in economy, an improvement in supply of materials and

goods, and a growth in trade. The number of cities at that time grew

40 1 Introduction

about 100 at the beginning of the second millennium up to 3,000 until the

14th century (Heinrich 1983).

With the growth of cities and trade, the requirement and the possibility

of constructing stone arch bridges also returned. A list of early medieval

stone arch bridges is given in Table 1-4.

Table 1-4. List of medieval arch bridges in Europe (Velflík 1921, Heinrich 1983

and Mehlhorn and Hoshino 2007)

City Bridge construction time River

Toledo Approximately 900 Tagus

Würzburg Probably 1133–1146 Main

Regensburg 1135–1146 Danube

Prague 1158–1172 Moldavia

London 1176–1209 Thames

Avignon 1178–1188 Rhone

Dresden 1179–1260 Elbe

more. Regensburg was, at that time, one of the biggest cities in the German

Empire and, besides Cologne, the second biggest trading centre. Several

central Europe trade roads crossed in Regensburg. The relations to the

South were of major importance since Venice and Genoa were the com-

mercial and maritime super powers of the western Mediterranean Sea and

had possessed excellent relations with the eastern Mediterranean Sea too.

Because the German Empire at that time included major parts of Italy,

Germany, and some parts of France, Regensburg was located very cen-

trally. In contrast, Cologne was more important for trade with England

(Ludwig and Schmidtchen 1992).

However, besides the central location of Regensburg in the German

Empire, Regensburg had a difficult binding to the sea. Shipping at that

time was of major importance, since transport on roads was the most ex-

pensive method. The cost of customs, escorts, damages on vehicles, over-

night stay, and food for animals had to be provided. Transport on roads

was about five times more expensive than transporting the goods on rivers

and about ten times more expensive than transports on sea (Ganshof

1991).

Furthermore, not only the costs were important, but uncertainties about

the time schedules were also considerable. In spring and in fall, fords were

often impassable and traders had to wait weeks to pass the rivers. For ex-

ample, Wilhelm the Conqueror had to wait three weeks in 1069 to pass the

river Aire on his way to York (Harrison 2004). Sometimes major rivers

1.7 History of Stone Arch Bridges 41

were even seen as invincible barriers, such as the Yangtze in China (Yi-Sheng

1978).

Therefore, Regensburg had to improve the attraction of road transport.

So, when in the year 1135 an unusual drought yielded to a very low river

level, the construction of the stone arch bridge was launched. The bridge

design was mainly based on historical Roman templates. However, in con-

trast to the original Roman bridges, the foundations were different, since

Roman concrete was forgotten in medieval ages. Therefore, the piers had

to be protected against water in a different way: islands were constructed

around the piers. The arches had a span between 10.4 and 16.7 m, but the is-

lands yielded to a greater constriction of water flow between the piers.

It has been assumed by historians that the experiences from the con-

struction of the Regensburg arch bridge spread in Europe. For example, for

the construction of the bridges in Prague and Dresden, knowledge from

Regensburg was probably used. Furthermore, in many different cities

bridge construction schools may have evolved after such important and

successful constructions. For example in France, the friars of bridge con-

struction were led by Saint Bénézet, the designer of the arch bridge in

Avignon. Although the existence of such an order could never be proven,

the construction of the arch bridge in Avignon was a milestone in the art of

bridge construction. The bridge consisted not only of an incredible span of

33 m, but also on a very slender vertex. In direct comparison to the arch

bridge in Regensburg, the arch bridge from Avignon looks much more

slender and graceful, although the bridge was constructed only 40 years

later. This excellent expression is also reached by the application of circu-

lar segments (Heinrich 1983, Brown 1994).

The application of circular segments actually became very popular

only in the Renaissance. Besides the success of the segmental arch, the

basket arch was applied more and more. Both yielded to more elegant

views, wider spans, and lower bridge access roads than using a half-

circular arch shape. A third shape, the ellipse shape, also became widely

known, especially due to Dürers publication about the construction of el-

lipses. However, the ellipse could not assert itself and the basket arch was

much more successful (Heinrich 1983).

During the 14th and 15th centuries, the basket arch and circular arch

segments were widely used. For example, in Italy during that time, many

famous bridges were constructed, such as the Ponte Vecchio in Florence.

This bridge constructed in 1341–1345 reached a span of 32 m by a ratio of

pier width to a span of only 1:6.5 (Heinrich 1983, Brown 1994). Only 200

years later, Ammanati designed another famous bridge in Florence: the

Ponte Santa Trinità. The three-span bridge built from 1567 to 1569 reaches

spans of 32 m. The shape of the arch also follows a basket arch, but the

major clue of the bridge is the ratio of the pier width to the span of 1:7.

42 1 Introduction

The bridge was so slender that the public could not believe the sufficient

load-bearing behaviour of the bridge for some time. As a final example for

the Italian bridge design art at that time, the Rialto Bridge in Venice, de-

signed by Da Ponte and constructed in 1588–1591, is mentioned. The

bridge has only a span of 28 m but the foundation of the bridge is extraor-

dinary: about 12,000 wooden piles were used (Heinrich 1983, Brown

1994).

Only with the butcher bridge in Nuremberg, built in 1597–1602, could

the German bridge designers reach the quality of the Italian bridge con-

structors again. The bridge reached a span of 34 m but shows many paral-

lels to the Rialto bridge in Venice and was probably therefore strongly in-

fluenced by Italian bridge designers.

About 150 years later, under the French bridge designer Perronet, the

construction of stone arch bridges reached its perfection. Perronet designed

the bridge in Neuilly and the Concorde Bridge in Paris. The bridge in

Neuilly built in 1768–1774 was a landmark in arch bridge design. The

bridge consisted of five arches with a maximum span of nearly 38 m. The

piers were extremely slender and had a width of only 4.2 m. This gives a

ratio of span-to-pier width of 1:9.23.

The development of such excellent bridges did not come from any-

where. Just like the Romans, the central organized French state at that time

had recognized the importance of a good road and bridge system. To pro-

vide that in 1716, a Corps des Ingéniurs des Ponts et Chaussée was

founded. In 1747 in Paris, the school “Ecole des Ponts et Chaussées” was

founded and Perronet was the head of this school for more than 47 years.

Besides his teaching, he strongly improved the arch bridge design as ex-

amples have shown. He discovered that the arch piers, need to take vertical

loads only or vertical loads and very limited horizontal loads if the adja-

cent arches are constructed at the same time. This would permit very slen-

der piers, as constructed at the bridge in Neuilly. Furthermore, he heavily

used slender basket arches and thus eliminated the bridge access road

(Heinrich 1983, Brown 1994, Jesberg 1996).

Figure 1-38 summarizes the development of arch bridges over time.

However, the last shapes already show the rise of new construction materi-

als: steel and reinforced concrete.

1.7 History of Stone Arch Bridges 43

of arch bridges. He investigated the development of bridges over certain

rivers in England from around 1540 to the middle of the 19th century (Ta-

bles 1-5 and 1-6). He furthermore tried to distinguish between arch bridges

and other bridges. In contrast to the aforementioned authors, Harrison as-

sumes that the first stone bridges in England were already there at the end

of the 11th century. That would be around 50 years before the bridge in

Regensburg. However, Harrison also mentions the problems in the defini-

tion of stone bridges. For example, stone bridges could have also been

bridges with stone piers and wooden superstructures or with stone para-

pets. According to Harrison, the oldest (arch) bridge built in Oxford was

ordered by Robert D’Oilliy at the end of the 11th century. A second stone

bridge was built in Winchester and is perhaps even older. The estimation

of the age and the construction type of the later bridge are, however, very

uncertain. Around 1086 there should have been several stone bridges in

England. At around 1100 the number of stone bridges increased and also

the historical indications became more trustworthy (Harrison 2004).

44 1 Introduction

Besides England, Harrison (2004) also states that stone bridges were

built at that time in France. He assumes such bridges over the Loire in

Blois and Tours before 1100. Further stone arch bridges were built in

Saumur in 1162, in Orleans in 1176 and in Beaugency in 1160–1182

(Harrison 2004).

In general, stone arch bridges were very common in Great Britain be-

tween the late Middle Ages and the end of the 19th century. In the 16th

century, stone arch bridges were the major part of the bridge stock. Some

examples illustrate that: the old Exe Bridge in Exeter, which was found by

excavations during 1960 and 1970, was probably built in the 12th century.

The Framwellgate Bridge was built around 1400. The span of the bridge

was exceeded only in the middle of the 19th century. A last example is the

Warkworth Bridge erected around 1380 with a span of 20 m. Ruddock

(1979) gives a good summary and description about constructions of arch

bridges in Great Britain and Ireland in the 18th century. Therefore, the list

of successful arch bridges in these countries could be further extended.

Many of the historical arch bridges perform very well even today. And al-

though age, increasing traffic, floods and storms attacking the bridges, still

today the major cause of destruction of historical arch bridges is an insuf-

ficient road width of the bridges (Harrison 2004).

Table 1-5. Development of the number of bridges over English rivers over time

(Harrison 2004)

River 1540 1765–75 1850

Avon (downstream from the Finford Bridge) 17 18 20

Great Ouse (from Claydon Brook to Ely) 17 24 36

Severn (from Montford Bridge) 10 10 16

Thames (from Lechlade) 17 23 36

Trent (from Stoke-on-Trent) 16 23 30

Ure und Ouse (from Bain Bridge) 10 12 16

Avon (Bristol) (downstream from Malmsbury) 13 18 21

Avon (Hants.) (from Salisbury) 7 10 11

Medway (from Ton Bridge) 8 10 12

Stour (Dorset) (from Blandford) 6 7 7

Tame (Staffs.) (from Water Orton) 6 9 9

Wear (from Stanhope) 9 12 15

1.7 History of Stone Arch Bridges 45

2004)

River Number Number Number Building

of of stone of material

bridges bridges wooden unknown

bridges

Avon (downstream from the Finford Bridge) 17 13 3 1

Great Ouse (from Claydon Brook to Ely) 17 8 2 7

Severn (from Montford Bridge) 10 9 1 0

Thames (from Lechlade) 17 7 9 1

Trent (from Stoke-on-Trent) 16 5 1 10

Ure und Ouse (from Bain Bridge) 10 6 2 2

Avon (Bristol) (downstream from 13 9 0 4

Malmsbury)

Avon (Hants.) (from Salisbury) 7 6 0 1

Medway (from Ton Bridge) 8 7 0 1

Stour (Dorset) (from Blandford) 6 6 0 5

Tame (Staffs.) (from Water Orton) 6 4 0 2

Wear (from Stanhope) 9 6 0 3

The oldest historical arch bridges in the state of Saxony, Germany origi-

nate from the 13th century, such as the Elster Bridge in Kürbitz and a bridge

in Plauen. The Hammer Bridge (1550–1576) and the Altväter Bridge

(around 1570) on the Mulde were constructed in the 16th century. The

state of Saxony experienced an economical boom period mainly due to the

silver mining at that time. The end of the silver mining and the 30 Years

War caused an economical decline and the end of the stone arch construc-

tion period. Only the inauguration of the Saxon elector and later Polish

king Friedrich August I. again yielded to an economical prosperity period

with further stone arch bridges. For example, in 1716–1719 in Grimma,

and in 1715–1717, the Pöppelmann Bridge and a bridge in Nossen over the

Mulde were constructed (Frenzel 2004).

During the origin of the Saxon railway net between 1846–1851, the

world’s biggest brick stone arch bridge, the Göltzschtal Bridge, was built.

Furthermore, the Syratal Bridge in Plauen, constructed in 1905 with a span

of 90 m, was a major landmark for stone arch bridges. It was (or still is)

the masonry stone arch bridge with the longest span worldwide (Frenzel

2004).

However, already at that time, the boundary conditions for bridge con-

struction changed increasingly. The Industrial Revolution during the 19th

century, with the introduction of new means of transport such as the rail-

way, yielded to an exponential growth of transported goods. This caused

completely new requirements for bridge constructions. Furthermore, the

scientific approach to solving engineering problems became more and

46 1 Introduction

dealing with bridge design (BMA 2008). For example, in 1809, Wiebeking

(1809) published material about a planned bridge. In 1847, Ardant, profes-

sor for construction art in Metz, published a book about the construction of

bridges including material about arch bridges. He furthermore developed a

device to test arch bridge models for bending and shear. Kurrer and Kah-

low (1998) show the device from Ardant. Based on these results and Na-

viers’ theory, Ardant was able to compute the deformation of the arch at

the crown with

1 K ⋅V (1-2)

y= ⋅ ⋅ l ⋅ h2

2 E⋅I

with y as deformation at crown, l as span, h as rise, V as sum of vertical

loads, EI as bending stiffness, and K as factor for the consideration of the

load distribution over the arch.

Schubert (1847/1848), the designer of the Göltzschtal Bridge, published

his own vault theory in 1847 and 1848, and Scheffler published his vault

theory in 1857 (Scheffler 1857).

After the 1920s, scientists started investigating deformations of model

arch bridges in more detail. The major goal was the avoidance of the com-

plicated numerical computation of static indeterminated arches. The tests

included an arch model, a device for the application of the load and a mi-

croscope for the observation of the arch deformations. The devices were

called deformers. There were different deformers developed, for example

one by Beggs (1927), Bühler (1927), and Magnel and Schaechterle

(Mörsch 1947). In Mörsch (1947), an illustration of the deformer by

Schaechterle applied on a fixed arch is shown.

The last paragraphs gave the impression that statical computations ar-

rived only very late in arch bridge design. In contrast, Fleckner (2003) as-

sumes that already around the year 1200 simple static computations were

carried out for vaults and piers in France. He proves that sufficient knowl-

edge about mathematics and mechanics was provided at that time, espe-

cially in cloister schools. Fleckner (2003) further shows that independent

from the different shapes and designs, all piers in churches experienced

pressure under normal loading conditions. The avoidance of tensile stresses

inside the piers is, according to Fleckner, satisfactory proof of the men-

tioned computation at the former times.

However, advanced statical computations were used for arch bridges us-

ing other materials such as reinforced concrete or steel.

1.8 Arch Bridges from Alternative Material 47

The oldest historical cast iron arch bridge was constructed over the Severn

Coalbrookdale in England in 1779 (Martínez 2004, Schaechterle et al.

1956, Brown 1994). The bridge consisted of five parallel arches with a

span of 30.5 m. The designer was Thomas Pritchard. Twenty years later,

Telford designed a cast iron arch bridge with a span of 40 m. In 1796 in

Germany, the first iron road arch bridge was built over the Striegauer Wa-

ter in Silesia. In the first half of the 19th century, several further such

bridges were built in Germany and England. The maximum span until the

middle of the 19th century was reached with 72 m in England (Martínez

2004).

In the middle of the 19th century, the material changed from cast iron to

wrought iron, allowing much larger spans. In 1884, Gustav Eiffel designed

the Garabit Viaduct in France with a span of 165 m. However, the applica-

tion of wrought iron lasted a rather short time. Already since 1860, steel

had became more and more popular. The Mississippi Bridge in St. Louis,

built from 1867 to 1874, was probably the first large steel arch bridge

worldwide with a span of 158.5 m. The bridge is a three-span bridge and

was designed as combined road and railway bridge. The bridge is now

named after the designer Eads (Martínez 2004).

In Germany, the Wuppertal Bridge in Müngsten reached 160 m in 1893,

and the Rhine Bridge Engers-Neuwied reached 188 m in 1918

(Schaechterle et al. 1956).

Over the next decades, steel arch bridges experienced a major develop-

ment. Using bending stiff slabs, arch bridges spans of up to 250 m were

reached, with arch frameworks up to 500 m span possible. International

examples are the Mälarseer Bridge in Stockholm (1935) with a 204 m

span, the Henry Hudson Bridge in New York with a 244 m span, or the

Sydney Harbour Bridge in Australia in 1930 with a 503 m span.

In 1977, the New Gorge River Bridge in Fayetteville was opened. The

bridge has the span of 518 m. The current record is owned by the Lunpu

arch bridge in Shanghai, China, with an incredible 550 m. The bridge was

opened in 2003. Furthermore, the latest developments are so-called net-

work arch bridges, which yield to extremely slender steel arch bridges

(Tveit 2007 and Graße and Tveit 2007).

48 1 Introduction

made of wood represents a problem. However, with the skilled connection

of single wood elements, wooden arch bridges become possible. One of

the first visualizations and indications of a wooden arch bridge can be

found at the Trajan pillar in Rome. The pillar was built in 110 A.D. and

shows a wooden arch bridge over the river Danube, which was probably

constructed during the Dacia wars (Steinbrecher 2006).

Leonardo da Vinci described a wooden arch bridge in the 15th century

(Ceraldi and Ermolli 2004). In 1540, Andrea Palladio described a wooden

arch bridge in Northern Italy (Fletcher and Snow 1976). In the 18th cen-

tury, several wooden arch bridges were constructed in England, such as the

Walton bridge over the river Thames (Ceraldi and Ermolli 2004). In the

18th century, also in France, the wooden arch Bridge Point de Choisy sur

la Seine was built (Fletcher and Snow 1976). Also in Germany, wooden

arch bridges were constructed (Holzer 2007).

Ivan Petrovic Kulibin designed a wooden arch bridge with a span of

300 m in the middle of the 18th century in Russia. The bridge was planned

for Saint Petersburg to span the river Neva. In December 1776, a loading

test on a model with scale 1:10 was carried out. Experts for evaluation of

the model were, for example, Leonhard Euler and Daniel Bernoulli. Unfor-

tunately, the bridge was never built (Bühler 2004).

In Japan, a historical wooden arch bridge has been considered as a most

important historical bridge: the Kinai-Kyo Bridge (Bühler 2004, Troyano

2003). Yang et al. (2007) describe the development of wooden arch

bridges in China.

The Colossus Bridge was the tallest wooden arch bridge ever built. It

was designed by Lewis Wernwag over the river Schuylkill at Fairmont in

Philadelphia in 1811. The bridge reached a span of 103.60 m. However, a

fire destroyed the bridge in 1838 (Bühler 2004, Troyano 2003 and Fletcher

and Snow 1976).

In Venice between 1933 and 1934, the wooden arch bridge Ponte dell’

Accademia was constructed as a temporary structure. However, the bridge

performed so well that the bridge remains functioning.

Modern wooden arch bridges show many advances compared to histori-

cal bridges. For example, parts of the structure are made of glued-

laminated timber and the roadway is made of concrete as shown at the

Crestawald bridge in Switzerland or the Wennerbridge in Austria (Bühler

2004). A last interesting example of a wooden arch bridge is the Leonardo

Bridge in Norway (Goldberg 2006).

1.8 Arch Bridges from Alternative Material 49

was only a pedestrian bridge. Only 30 years later, in 1904, Hennebique de-

signed the Risorgimento Bridge in Rome with a span over 100 m (Martínez

2004). Until the end of the 19th century, concrete arch bridges were

mainly built with a Spangenberg’s boldness number up to 700. Spangen-

bergs number is computed by the span to the square divided by the rise.

The Risorgimento Bridge in Rome reached a Spangenberg’s number of

1,000, manifesting the change from the arch to the beam (Pauser 2002).

Maillard and Freyssinet further developed the reinforced concrete arch.

For example, Freyssinet constructed an arch bridge in Villeneuve-sur-Lot

with a span of 100 m in 1910. In 1925, he constructed an arch bridge in

Plougastel consisting of three arches with a span of 180 m each. At that

time, the formwork crystallized as problem.

From 1938 to 1942, the Sando Bridge (with a span of 264 m) was con-

structed. Until the finishing of the Arrabida Bridge in Porto, the Sando

Bridge remained the largest arch bridge worldwide. The Arrabida Bridge

was opened in 1963. The Arrabida arch reaches a span of 270 m. In the

same year, the Gladesville Bridge in Australia was opened with a span of

305 m. In 1979 in Croatia, the Krk arch bridge with a span of 390 m, was

finished. Currently, the largest concrete arch bridge is the Wanxian Bridge

over the river Jangtse in China, with a span of 420 m. In Germany in re-

cent years, the Kylltal Bridge was opened and, although it does not reach

the span of the Wanxian Bridge, the bridge is a good example of a concrete

arch bridge.

The tenders for the Millau Viaduct in France also included a reinforced

concrete arch bridge with a span of 600 m. Kamisakoda et al. (2004) have

also discussed concrete arch bridges with spans up to 600 m. Čandrlić

et al. (2004) have published numerical investigations for a reactive powder

concrete (RPC) arch bridge with a span of 1,000 m. Martínez (2004) sug-

gests an economical usage of concrete arch bridges with a maximum span

of 1,000 m and an normal span of 500 m.

Arch bridges built with modern construction materials such as steel and

concrete are still very competitive bridges. New materials or construction

technologies, such as concrete-filled steel tubular arch bridges, may even

increase the applicability of arch bridges and the competitiveness such as

recent examples in China show (Chen et al. 2004, Martínez 2004, Chen

2007).

50 1 Introduction

Since the history of stone arch bridges is strongly related to the development

of new building materials, as stated before, the question arises how many

stone arch bridges still exist about 150 years after the massive implementa-

tion of new building materials such as reinforced concrete or steel.

In Germany, currently about 120,000 bridges exist (Prüfingenieur

2004). Current estimations assume a value of 80 billion Euro for all the

120,000 bridges in Germany. Weber (1999) has estimated the value of all

the German railway bridges with a span lower than 20 m with about 66 bil-

lion DM in 1993 prices. The reconstruction cost of the entire railway

bridge stock would probably exceed 26 billion Euro according to Marx

et al. (2006).

Although many of these bridges were constructed in recent decades,

stone arch bridges still contribute significantly to this bridge stock. In

some regions, they still represent the majority of bridges. For example, in

the region of the road department of Zwickau (Saxony, Germany), about

1/3 of all bridges are historical stone/concrete arch bridges that were con-

structed at the end of the 19th century (Bothe et al. 2004). The overall

number of road bridges in the German federal state of Saxony was esti-

mated at the beginning of 1990 as approximately 4,000 (Bartuschka 1995).

About 32% of this bridge stock was estimated as stone arch bridges (Bar-

tuschka 1995). Purtak (2004) assumes several thousand stone arch bridges

in Saxony.

Schmitt (2004) presumes that about 1/3 of all railway bridges with a

span between 10 and 20 m are historical stone arch bridges in Germany.

The German railway organization (DB AG) has estimated the overall

number of stone arch bridges in the German railway system at about

35,000. That would represent about 40% of all bridges and culverts, and

about 30% of bridges with a minimum span of 2 m (without the culverts)

(Orbán 2004).

According to Marx et al. (2006), the German railway currently uses

29,200 railway bridges, 860 road bridges, and another 1,300 bridges for

pipes, signal transfer, and other purposes. Most of the bridges have a span

less than 30 m (96%). They assume that historical arch and vault bridges

made up 28% of the entire bridge stock. About 25% of the bridges are roll-

ing steel beams in concrete, 24% are steel bridges, 18% are reinforced con-

crete bridges, and 4% are prestressed bridges. According to them, most

bridges were built between 1900 and 1920, and a second peak was between

1970 and 1995. The average bridge age is 70 years. This is only due to the

fact that since the reunification of Germany more than 270 new bridges have

been constructed. Before that, the average bridge age was 80 years old and

bridges with an age of 170 years are still in use (Marx et al. 2006).

1.9 Number of Arch Bridges 51

wooden bridges until approximately 1860. After the introduction of weld-

ing steel and the construction of rolling steel beams in concrete, the num-

ber of new stone arch bridges dropped (Fig. 1-39). Only after World War

II, several arch bridges were rebuilt, however mainly from concrete. Since

then, the construction of new stone arch bridges rests (Weber 1999).

excluding wooden bridge structures. Wooden railway bridges were not permitted

after 1865 based on the technical requirements of the association of German rail-

way administration. Wooden railway bridges were built until 1860 as permanent

structures with a span up to 40 m. In many other countries, wooden structures

made a significant part of the bridge stock (Weber 1999).

about 150,000 (Woodward et al. 1999). The number of stone arch bridges

in the British railway system ranges in certain publications between 20,000

(UIC 2005, Orbán 2004, Melbourne et al. 2004,) and 40,000 (Choo et al.

1991). Murray (2004) assumes an overall number of about 40,000 arch

bridges in Great Britain. Harvey et al. (2007) have estimated the overall

British railway bridge stock at about 70,000. Smith (2003) estimates

about 2,400 arch bridges, mainly brick and stone arch bridges owned and

52 1 Introduction

(1999) have estimated that from 13,000 bridges belonging to highway and

long-distance roads, about 0.7% are stone arch bridges.

Brencich and Colla (2002) estimate the number of stone arch bridges

with a span greater than 8 m, and constructed in the second part of the 19

century in the Italian railway system, as at least 7,000. Cavicchi and

Gambarotta (2004) number the quantity of masonry arch bridges in the

Italian railway with a span greater than 2 m at 12,000, whereas 80% of this

bridge stock has a span smaller than 5 m. Harvey et al. (2007) have esti-

mated the overall Italian railway bridge stock at about 180,000.

In 1985, Spain launched a systematic registration of the bridge stock of

the Spanish highway system. Until 1996, more than 8,000 bridges were re-

corded. About 2,824 of these bridges are masonry bridges, which repre-

sents about one third. However, especially for short spans, masonry bridges

show a much higher contribution: up to 70%. In contrast with wider spans,

the ratio of masonry bridges drops to less than 10% (Angeles Yáñez and

Alonso 1996).

In France, about 21,000 bridges belong to the highway and long-

distance roads system, and about 24% of them are masonry bridges. For

comparison purposes, the France railway system includes 100,000

bridges (Harvey et al. 2007). In Spain, masonry bridges represent about

8%, and in Slovenia about 6%, of the highway bridge stock (Woodward

et al. 1999). In Norway, masonry bridges make only a minor contribution

to the highway bridge stock of about 1.3% (Woodward et al. 1999). The

THC (2006) reports about stone arch bridges in Ireland.

Bién and Kamiński (2004) assume approximately 1,000 railway stone

arch bridges and about 2,000 road stone arch bridges in Poland. However,

they have counted only bridges with a minimum span of 3 m. If culverts

are included, then the overall number increases to more than 20,000.

Radomski (1996), in contrast, considers the contribution of masonry

bridges to the Polish road bridge stock as negligible and does not re-

cord it statistically.

Karaveziroglou-Weber et al. (1998) estimate the number of stone arch

bridges for central Greece as 300. Turer (2008) estimates the overall num-

ber of bridges in Turkey at about 5,000 and 100–200 are stone arch

bridges. Dogangün and Ural (2007) also describe certain Anatolian stone

arch bridges in some other regions of Turkey (Ural and Dogangün 2007).

The overall road bridge number in a Swiss Canton was given as 667 for

2,126 km of road distance. Masonry arch bridges make up 22%, and ma-

sonry-concrete arch bridges contribute 37% to the mentioned bridge stock

(Brühwiler 2008).

The United States probably has a total bridge stock of more than

600,000 bridges (Gaal 2004, Dunker 1993). However, masonry arch bridges

1.9 Number of Arch Bridges 53

contribute only a minor stock with less than 1,000 (Boothby and Roise

1995). Perhaps, because this low number increases the efforts to retain the

bridges, several descriptions and collections about stone arch bridges in the

United States can be found (SABM 2006, Evans 2006, Forbes 2006,

HBMW 2006). In Virginia, US, about 187 historical arch bridges built

from concrete and masonry are recorded (Miller et al. 2000). Senker (2007)

gives details about the number of stone arch bridges in Pennsylvania.

Generally, in the United States, good records about the bridges was intro-

duced after the collapse of the Silver Bridge at Pleasant Point, West Vir-

gina in 1967. The collapse yielded to the introduction of the Federal Highway

Act in 1968, which requires the recording and documentation of bridges.

However, as mentioned before, the number of arch bridges as part of the

highway system is extremely low. Even reinforced concrete arch bridges

contribute only minor to the bridge stock with about 0.2% (Gaal 2004,

Dunker 1993).

All types of arch bridges contribute according to unofficial statistics,

with 70% to the overall bridge stock of China (Dawen and Jinxiang 2004).

Ou and Chen (2007) describe stone arch bridges in a region of China. Even

in recent times, stone arch bridges have been constructed in China such as

the New Danhe Bridge that opened in 2000 (Chen 2007). The overall

number of road bridges in China is 533,618 for 3,457,000 km road (Yan

and Shao 2008). The first stone arch bridge in Korea was erected in 413,

the latest still existing stone arch bridge originates from the year 760

(Hong et al. 2008). Hong et al. (2007) describe certain properties of his-

torical Korean stone arch bridges.

The Indian railway has 119,724 bridges–of which 20,967 are arch

bridges according to Gupta (2008). Damage on Indian stone arch bridges is

also reported (Bridge 2006). Mathur et al. (2006) describe the assessment

and strengthening of arch bridges in the Indian railway system.

Shrestha and Chen (2007) mention stone arch bridges in Nepal. Exam-

ples of arch bridges in Russia, for example in Saint Petersburg, are given

in The Heritage Council (THC 2006). Seiler (2006) reports about masonry

railway bridges in Eritrea (Africa). Paredes et al. (2007) mention stone

arch bridges in Colombia, South America.

In general, it is difficult in all the statistics to figure out the meaning of

masonry bridges, arch bridges, and stone arch bridges, and so on. This, to a

certain extent, limits the comparability of all these numbers.

Therefore, a systematic analysis of the arch/vault bridge stock in the in-

ternational railway system was carried out by the International Union of

Railways. Thirteen railway organizations contributed and over 200,000

railway arch bridges were counted (UIC 2005). Harvey has estimated the

number of arch bridges in Europe even much higher. Based on the UIC

(2005) report, stone arch bridges contribute up to 60% of the bridge stock of

54 1 Introduction

the railway organizations joining the study. A more detailed summary of the

results of the investigation is shown in Table 1-7 and Figs. 1-40 and 1-41.

It is interesting to compare the data with a former study published by

Weber (1999). He also has investigated the amount of stone arch bridges in

the European railway stock. The results are summarized in Table 1-8.

Table 1-7. Number of stone arch bridges in different European railway organiza-

tions (UIC 2005)

Railway organization Number of Number of Ratio of stone Ratio of stone

stone arch stone arch arch bridges arch bridges

bridges and bridges on all bridges on all bridges

culverts and culverts without

culverts

French (SNCF) 78,0001 18,060 76.8 43.5

Italian (RFI) 56,888 94.5

British (NR) 17,867 16,5001 46.9

Portuguese (REFER) 11,746 874 89.8 39.6

German (DB) 35,0001 8,653 38.9 27.5

Spanish (RENFE) 3,144 49.3

Czech (CD) 4,858 2,391 18.9 35.8

1

Estimated number

Fig. 1-40. Proportion of arch bridges on the bridge stock of several European

railway organizations (UIC 2005)

1.9 Number of Arch Bridges 55

Fig. 1-41. Distribution of certain properties of the historical stone arch bridges in

the European railway bridge stock (UIC 2005)

Table 1-8. Number of stone arch bridges in different European railway organiza-

tions according to Weber (1999)

Railway organizationOverall Number Bridge Number Ratio in Oldest

railway of railway density2 of arch percent 3 arch

distance bridges1 bridges bridge

in km from

Belgium (SNCB/NMBS) 3,432 3,400 10 600 18 1845

British (BR) 16,528 26,240 16 13,000 50 1825

Bulgarian (BDŽ) 4,299 982 2 62 6 1867

Danish (DSB) 2,344 1,500 6 135 9 1853

German (DB) 4,087 32,017 8 9,146 29 1837

Finnish (VR) 5,874 1,905 3 60 3 1861

French (SNCF) 32,731 28,259 9 13,167 47 1840

Greek (CH, OSE) 2,484 21,000 8 710 34 1883

Italian (FS) 16,112 59,473 37 37,400 63 1850

Irish (CIE) 1,944 2,752 14 1,484 54 1839

Yugoslavian (JŽ) 2,770 619 22 1874

Luxemburg (CFL) 275 282 10 149 53 1859

56 1 Introduction

Norwegian (NSB) 4,027 2,700 7 311 12 1888

Austrian (ÖBB) 5,605 5,048 9 1,200 24 1838

Polish (PKP) 25,254 8,500 3 1,020 12 1842

Portuguese (CP) 3,054 1,928 6 883 46 1875

Rhaetian (RhB) 375 489 13 931 1888

Rumanian (CFR) 11,430 4,067 4 240 6 1859

Swedish (SJ) 9,846 3,500 4 100 3 1857

Swiss (SBB)* 2,985 5,267 18 914 17 1847

Spanish (RENFE) 13,041 6,371 5 3,205 50 1860

Czechoslovakia (ČSD) 13,100 9,411 7 3,213 34 1845

Hungarian (MAV) 7,605 2,375 3 278 12 1845

1

Crossed by trains

2

Number of bridges per 10 km railway distance

3

Ratio of arch bridges on the overall bridge stock

* The value of 1,000 arch bridges at the Swiss Railway is also mentioned by

Berset (2005)

All these numbers give the impression that stone and brick arch bridges

performed very well over the last decades and centuries, otherwise they

would not still contribute heavily to European and worldwide bridge stock.

According to Jackson (2004), arch bridges require the lowest maintenance

costs of all bridge types. The high maintenance costs currently are mainly

caused by the high age of the bridges. The low lifetime costs may be one

reason why new brick and stone arch bridges were constructed in Great

Britain recently (Wallsgrove 1995). A new stone arch bridge has also been

erected recently in Portugal (Arêde et al. 2007).

However, besides the maintenance costs and limitations of the historic

bridges in fulfilling modern serviceability, historic bridges also have to

bear heavily on the change in live loads over the last century and the last

few decades. Therefore, it is of great interest not only to understand the

modelling of live loads as discussed in detail in Chapter 2, but also how to

limit and control the live load.

References

Angeles Yáñez M & Alonso AJ (1996) The actual state of the bridge management

system in the state national highway network of Spain. Recent Advances in

Bridge Engineering. Evaluation, Management and Repair. Proceedings of the

US-Europe Workshop on Bridge Engineering, organized by the Technical

University of Catalonia and the Iowa State University, JR Casas, FW Klaiber

& AR Marí (Eds). Barcelona, 15–17 July 1996, International Center for

Numerical Methods in Engineering CIMNE, pp 99–114

References 57

Construction der Sprengwerke von grosser Spannweite mit besonderer

Beziehung auf Dach- und Brücken-Constructionen aus geraden Theilen, aus

Bögen, oder aus der Verbindung beider; für praktische Baumeister, so wie für

Vorträge über Ingenieur-Mechanik; mit einem Atlas von 28 Tafeln und in den

Text gedruckten Holzschnitten. Auf Befehl des Französischen

Kriegsministeriums gedruckte Abhandlung. Deutsch hrsg. von Aug[ust] von

Kaven. Hannover (Hahn)

Arêde A, Costa P, Costa A, Costa C & Noites L (2007) Monitoring and testing of

a new stone arch bridge in Vila Fria, Portugal. Proceedings of the Arch 07 –

5th International Conference on Arch Bridges, PB Lourenço, D Oliveira &

V Portela (Eds), 12–14 September 2007, Madeira, pp 323–330

Bartuschka P (1995) Instandsetzungsmaßnahmen an hochwertiger Bausubstanz

von Gewölbebrücken. Diplomarbeit. Technische Universität Dresden, Institut

für Tragwerke und Baustoffe

Becqué RF (1983) De brug in Sociaal en historisch perspectief. In: Over Bruggen,

H de Jong (Ed), Delft University Press, The Netherland, pp 221–272

Beggs, GE (1927) Der Gebrauch von Modellen bei der Lösung von statisch

unbestimmten Systemen. Beton und Eisen, Heft 16, pp 300–304

Bennet D (1999) The Creation of Bridges – From Vision to Reality – The

Ultimate Challenge of Architecture, Design and Distance. Arum Press Ltd.,

London

Berset T (2005) Plastische Berechnung von Naturstein-Gewölbebrücken.

Kolloquium am Institut für Baustatik und Konstruktionen, ETH Zürich, 28.

Juni 2005

Bién J & Kamiński T (2004) Masonry Arch Bridges in Poland. Arch Bridges IV –

Advances in Assessment, Structural Design and Construction, P Roca &

C Molins (Eds), CIMNE, Barcelona, pp 183–191

Birkhoff GD (1933) Aesthetic Measure. Harvard University Press, Cambridge,

MA

BMA (2008) Bibliotheca Mechanico-Architectonica. www.bma.arch.unige.it

Boothby TE & Roise CK (1995) An overview of masonry arch bridges in the

USA. Arch Bridges, C. Melbourne (Ed), Proceedings of the 1st International

Conference on Arch Bridges, Bolton 1995, Thomas Telford, London, pp 11–18

Bothe E, Henning J, Curbach M, Bösche T & Proske D (2004) Nichtlineare

Berechnung alter Bogenbrücken auf der Grundlage neuer Vorschriften. Beton-

und Stahlbetonbau 99, Heft 4, pp 289–294

Brencich A & Colla C (2002) The influence of construction technology on the

mechanics of masonry railway bridges, Railway Engineering 2002, 5th

International Conference, 3–4 July 2002, London

Bridges (2006) http://www.iitk.ac.in/nicee/EQ_Reports/Bhuj/bridge1.htm

Brown DJ (1994) Brücken – Kühne Konstruktionen über Flüsse, Täler, Meere.

Callwey Verlag München

Brühwiler E (2008) Rational approach for the management of medium size bridge

stock. Bridge Maintenance, Safety, Management, Health Monitoring and

Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis Group,

London, pp 2642–2649

58 1 Introduction

Beton und Eisen, Heft 16, pp 304–306

Bühler D (2004) Brückenbau. Deutsche Verlags-Anstalt GmbH, München

Čandrlić V, Radić J & Gukov I (2004) Research of concrete arch bridges up to

1000 m span. Arch Bridges IV – Advances in Assessment, Structural

Design and Construction, P Roca & C Molings (Eds), CIMNE, Barcelona,

pp 538–547

Cavicchi A & Gambarotta L (2004) Upper bound limit analysis of multispan

masonry bridges including arch-fill interaction. Arch Bridges IV – Advances

in Assessment, Structural Design and Construction, P Roca & C Molins (Eds),

Barcelona, pp 302–311

Ceraldi C & Ermolli ER (2004) Timber arch bridges: a design by Leonardo. Arch

Bridges IV – Advances in Assessment, Structural Design and Construction,

P Roca & C Molings (Eds), CIMNE, Barcelona, pp 69–78

Chandler HW & Chandler CM (1995) The analysis of skew arches using shell

theory. Arch Bridges, C Melbourne (Ed)_, Proceedings of the 1st

International Conference on Arch Bridges, Bolton 1995, Thomas Telford,

London 1995, pp 195–204

Chen B (2007) An overview of concrete and CFST arch bridges in China. Arch’07

– 5th International Conference on Arch Bridges. Madeira, pp 29–44

Chen B (2008) Arch bridges. College of Civil Engineering, Fuzhou University

China, www.arch-bridges.cn

Chen B, Chen Y & Hikosaka H (2004) The application of concrete filled steel

tubular arch bridges and study on ultimate load-carrying capacity. Arch

Bridges IV – Advances in Assessment, Structural Design and Construction,

P Roca & C Molings (Eds), CIMNE, Barcelona, pp 38–52

Choo BS & Gong NG (1995) Effect of skew on the strength of masonry arch

bridges. Arch Bridges, C Melbourne (Ed), Proceedings of the 1st International

Conference on Arch Bridges, Bolton 1995, Thomas Telford, London 1995,

pp 205–214

Choo BS, Coutie MG & Gong NG (1991) Finite-element analysis of masonry arch

bridges using tapered elements. Proceedings of the Institution of Civil

Engineers, Part 2, December 1991, pp 755–770

Corradi M (1998) Empirical methods for the construction of masonry arch bridges

in the 19th century. Arch Bridges: history, analysis, assessment, maintenance

and repair. Proceedings of the second international arch bridge conference,

A Sinopoli (Ed), Venice 6–9 October 1998, Balkema Rotterdam, pp 25–36

Czechowski A (2001) Structural Reliability Assurance – A Process Approach.

Safety, Risk, Reliability – Trends in Engineering, Malta

Dawen P & Jinxiang H (2004) Development, actuality and strengthening of

double curved arch bridges in China. Arch Bridges IV – Advances in

Assessment, Structural Design and Construction, P Roca & C Molins (Eds.),

Seite 444–450

Diamantidis D (2001) Assessment of Existing Structures. Joint Committee of

Structural Safety JCSS, Rilem Publications S.A.R.L., The publishing

Company of Rilem

References 59

78, pp 63–66

Ding D (1993) Antike und moderne chinesische Brücken. Beton- und

Stahlbetonbau 88, pp 289–296

Dogangün A & Ural A (2007) Characteristics of Anatolian stone arch bridges and

a case study for Malabadi Bridge. Proceedings of the Arch 07–5th

International Conference on Arch Bridges, PB Lourenço, D Oliveira &

V Portela (Eds), 12–14 September 2007, Madeira, pp 179–186

Drei A & Fontana A (2001) Load carrying capacity of multiple-leaf masonry

arches. Historical Constructions, PB Lourenço & P Roca (Eds), Guimarães,

2001, pp 547–555

Dunker KF (1993) Why America's bridges are crumbling, Scientific American,

266, No. 3, March, pp18–25

Ebeling W & Schweitzer F (2002) Zwischen Ordnung und Chaos. Komplexität

und Ästhetik aus physikalischer Sicht. In: Gegenworte, Zeitschrift für den

Disput über Wissen. Berlin-Brandenburgische Akademie der Wissenschaften,

Heft “Wissenschaft und Kunst”, pp 46–49

Evans GR (2006) The Clements Stone Arch Bridge – A Preservation Challenge.

http://home.comcast.net/~g.k.evans/

Fanning PJ, Salomini VA & Sloan DT (2003) Nonlinear modelling of a multi-

span arch bridge under service load conditions. Extending the Life of Bridges,

Concrete and Composites Buildings, Masonry and Civil Structures, MC Forde

(Ed), The Commonwealth Institute, London, 1st–3rd July 2003

FAZ (2005) Frankfurter Allgemeine Zeitung. Kartographie der Sehnsucht. 14th

August, Nr 32, p V 1

Fleckner S (2003) Gotische Kathedralen – Statische Berechnungen. Langfassung

eines Artikels im Bauingenieur 1/2003

Fletcher R & Snow JP (1976) A history of the development of wooden bridges.

Paper No. 1864, American Society of Civil Engineers, Transactions. ASCE

Historical Publications Nr. 4, New York

Forbes A (2006) Rehabilitation of the Keeseville Stone Arch Bridge,

http://www.dot.state.ny.us/design/css/files/2001r07.pdf

Frenzel M (2004) Untersuchung zur Quertragfähigkeit von Bogenbrücken aus

Natursteinmauerwerk. Diplomarbeit Technische Universität Dresden

Gaal GCM (2004) Prediction of Deterioration of Concrete Bridges. Dissertation,

TU Delft

Ganshof, FL (1991) Das Hochmittelalter. Propyläen Weltgeschichte, Band 5,

Propyläen Verlag, Berlin

Garbrecht G (1995) Meisterwerke antiker Hydrotechnik. Einblicke in die

Wissenschaft. B.G. Teubner Verlagsgesellschaft. Stuttgart

Gazzola P (1963) Ponti romani. Contributo ad un indice sistematico con studio

critico bibliografico, Ponti romani 2, Florence

Gilbert M & Melbourne C (1995) Analysis of multi-ring brickwork arch bridges.

Arch Bridges, C Melbourne (Ed), Proceedings of the 1st International

Conference on Arch Bridges, Bolton 1995, Thomas Telford, London 1995,

pp 225–238

60 1 Introduction

Bridges: History, Analysis, Assessment, Maintenance And Repair.

Proceedings of the Second International Arch Bridge Conference, A Sinopoli

(Ed), Venice 6.–9. October, Balkema Rotterdam, pp 109–118

Goldberg D (2006) Spanning the Globe – A user's guide to the new golden age of

bridges. http://www.wired.com/wired/archive/13.01/bridge.html

Goyet J (2001) Basic Considerations on Risk-Based Inspection for Offshore

Structures. Safety, Risk, Reliability – Trends in Engineering, Malta

Graf B (2005) Bridges that Changed the World, Prestel Publishing, Munich

Graße W & Tveit P (2007) Netzwerkbogenbrücken – Geschichte, Tragverhalten

und ausgeführte Beispiele. 17. Dresdner Brückenbausymposium, 13. März

2007, Dresden, pp 183–198

Griefe NF (2006) Bridge Case Study: Forteviot Bridge over the River Earn.

http://www.trp.dundee.ac.uk/research/glossary/bridge.html

Gupta RK (2008) Judging suitability of arch bridges for higher axle loading by

load testing. Bridge Maintenance, Safety, Management, Health Monitoring

and Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis Group,

London, pp 468–475

Harrison D (2004) The Bridges in Medieval England – Transport and Society

1400–1800. Clarendon Press, Oxford

Harvey B (2006) Fun Arches

http//billharvey.typepad.com/photos/fun_arches/index.html

Harvey B, Ross K & Orban Z (2007) A simple, first filter assessment for arch

bridges. Proceedings of the Arch 07 – 5th International Conference on Arch

Bridges, PB Lourenço, D Oliveira & V Portela (Eds), 12–14 September 2007,

Madeira. pp 275–279

Haser H & Kaschner R (1994) Spezielle Probleme bei Brückenbauwerken in den

neuen Bundesländern. Berichte der Bundesanstalt für Straßenwesen, Verlag

für neue Wissenschaft, Bergisch Gladbach

HBMW (2006) Historic Bridges of the Midwest

http://bridges.midwestplaces.com/ar/pulaski/lakeshore/

Heinrich B (1983) Brücken – Vom Balken zum Bogen. Rowohlt Taschenbuch

Verlag GmbH, Hamburg

Hodgson J (1996) The Behaviour of Skewed Masonry Arch Bridges. PhD Thesis,

University of Salford, 1996

Holzer SM (2007) Der Bogen im Dach. Bautechnik 84, Heft 2, pp 130–146

Hong NK, Koh H-M & Hong SG (2008) Structural safety of historical stone arch

bridges in Korea. Bridge Maintenance, Safety, Management, Health

Monitoring and Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis

Group, London, pp 757–764

Hong NK, Koh HM, Hong SG, Bae BS & Yoon WK (2007) Characteristics of

historical arch bridges in Korea. Proceedings of the Arch 07 – 5th

International Conference on Arch Bridges, PB Lourenço, D Oliveira & V

Portela (Eds), 12–14 September 2007, Madeira, pp 187–192

Huges TG & Blackler MJ (1997) A review of the UK masonry arch assessment

methods. Proceedings of Institution Civil Engineering Structures and

Buildings, 122, August, pp 305–315

References 61

restoration of architectural heritage. International Scientific Committee for

Analysis and Restoration of Structures of Architectural Heritage, Draft

Version No 11, September 2001

Jackson P (2004) Highways Agency BD on New Masonary Arch Bridges. 27th

February 2004. http://www.smart.salford.ac.uk/February_2004_Meeting.php

Janberg N (2008) International Database and Gallery of Structures. N Janberg

Internet content services, www.structurae.net

Jensen JS, Plos M, Casas JR, Cremona C, Karoumi R & Melbourne C (2008)

Guideline for load and resistance assessment of existing European railway

bridges. Bridge Maintenance, Safety, Management, Health Monitoring and

Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis Group,

London, pp 3658–3665

Jesberg P (1996) Die Geschichte der Ingenieurbaukunst. Deutsche Verlagsanstalt,

Stuttgart

Jurecka C (1979) Brücken – Historische Entwicklung, Faszination der Technik.

Verlag Anton Schrool & Co, Wien

Jurecka C (1979) Brücken. Historische Entwicklung, Faszination der Technik.

Verlag Anton Schroll Co., Wien, Wien-München

Kamisakoda K, Nakamura H & Nakamura A (2004) Studies on long-span

concrete arch bridge for construction at Ikarajima in Japan. Arch Bridges IV –

Advances in Assessment, Structural Design and Construction, P Roca & C

Molings (Eds), CIMNE, Barcelona, pp 548–558

Karaveziroglou-Weber M, Stavrakakis E & Karayianni E (1998) Damage of

existing stone bridges in Greece. Arch Bridges: History, Analysis,

Assessment, Maintenance and Repair. Proceedings of the Second

International Arch Bridge Conference, A Sinopoli (Ed), Venice 6.–9. October

1998, Balkema Rotterdam, pp 361–367

Klein W (2008) Ist Schönheit messbar? 12. Berliner Kolloquium der Gottlieb

Daimler- und Karl Benz Stiftung, 7. Mai 2008, Konrad Adenauer Stiftung,

Berlin

Koch W (1998) Baustilkunde – Das Standardwerk der europäischen Baukunst von

der Antike bis zur Gegenwart. Band 1 und 2. Faktum Lexikon Institut –

Bassermann Verlag, Bertelsmann Lexikon Verlag, Gütersloh

Kurrer K-E & Kahlow A (1998) Arch and vault from 1800 to 1864. Arch Bridges:

History, Analysis, Assessment, Maintenance and Repair. Proceedings of the

Second International Arch Bridge Conference, A Sinopoli (Ed), Venice 6.–9.

October, Balkema Rotterdam, pp 37–42

Kurrer K-E (2002) Geschichte der Baustatik. Ernst und Sohn Verlag für

Architektur und technische Wissenschaften GmbH, Berlin

Leliavsky S (1982) Arches and short span bridges. Design Textbook in Civil

Engineering: Volume VII, Chapman and Hall, London

Leonhardt F (1982) Brücken. Deutsche Verlagsanstalt. Stuttgart

Lourenço PB (2002) Computations on historic masonry structures, Structure

Engineering Material, 4, pp 301–319

Ludwig K-H & Schmidtchen V (1992) Metalle und Macht – 1000 bis 1600 –

Propyläen Technikgeschichte, Band 2, Propyläen Verlag Berlin

62 1 Introduction

Martínez SPF (2004) Arches: evolution and future trends. Arch Bridges IV –

Advances in Assessment, Structural Design and Construction, P Roca &

C Molings (Eds), CIMNE, Barcelona, pp 11–25

Mathur M, Kumar SV & Singh K (2006) Assessment and retrofitting of arch

bridges

Mehlhorn G & Hoshino M (2007) Handbuch Brücken, Entwerfen, Konstruieren,

Berechnen, Bauen und Erhalten, G Mehlhorn (Ed), Springer, Heidelberg,

Berlin

Melbourne C & Tao H (1995) The behaviour of open spandrel masonry arch

bridges. Arch Bridges, C Melbourne (Ed), Proceedings of the 1st International

Conference on Arch Bridges, Bolton, Thomas Telford, London, pp 239–244

Melbourne C & Tao H (1998) The behaviour of open spandrel brickwork arch

bridges. Arch Bridges: History, Analysis, Assessment, Maintenance and

Repair. Proceedings of the Second International Arch Bridge Conference,

A Sinopoli (Ed), Venice 6.–9. October 1998, Balkema Rotterdam, pp 263–269

Melbourne C & Tao H (2004) The behaviour of open spandrel masonry arch

bridges. Arch Bridges IV – Advances in Assessment, Structural Design and

Construction, P Roca & C Molins (Eds), Barcelona, pp 125–133

Melbourne C (1998) The collapse behaviour of a multi-span skewed brickwork

arch bridge. Arch Bridges: History, Analysis, Assessment, Maintenance and

Repair. Proceedings of the Second International Arch Bridge Conference,

A Sinopoli (Ed), Venice 6.–9. October, Balkema Rotterdam, pp 289–294

Melbourne C, Tomor AK & Wang J (2004) Cyclic load capacity and endurance

limit of multi-ring masonry arches. Arch Bridges IV – Advances in

Assessment, Structural Design and Construction, P Roca & C Molins (Eds),

CIMNE, Barcelona, pp 375–384

Miller AB, Clark KM & Grimes MC (2000) Final Report: A Survey of Masonry

and Concrete Arch Bridges in Virginia, Virginia Department of Transportation

Molins C & Roca P (1998) Load capacity of multi-arch masonry bridges. Arch

Bridges: History, Analysis, Assessment, Maintenance And Repair.

Proceedings of the Second International Arch Bridge Conference, A Sinopoli

(Ed), Venice 6.–9. October, Balkema Rotterdam, pp 213–222

Mori Y & Kato T (2003) Practical method of reliability-based condition

assessment of existing structures. Applications of Statistics and Probability in

Civil Engineering, Der Kiureghian, Madanat & Pestana (Eds), Millpress,

Rotterdam, pp 613–620

Mori Y & Nonaka M (2001) LRFD for assessment of deteriorating existing

structures, Structure Safety, 23, pp 297–313

Mörsch E (1947) Die Statik der Gewölbe und Rahmen, Teil B, Verlag Konrad

Wittwer Stuttgart

Mörsch E (1999) Gewölbte Brücken. Betonkalender 2000, Part 2, Verlag Ernst

und Sohn, Berlin, pp 1–53

Murray M (2004) CIRIA Research Project RP692 Masonry and Brick Arch

Bridges. Condition appraisal and remedial treatment.

http://www.smart.salford.ac.uk/February_2004_Meeting.php

NN (2005) Minzhu Bridge:

http://bridge.tongji.edu.cn/chinabridge/arch/133-2.HTM

References 63

Orbán Z (2004) Assessment, reliability and maintenance of masonry arch railway

bridges in Europe. Arch Bridges IV – Advances in Assessment, Structural

Design and Construction, P Roca & C Molins (Eds), Barcelona 2004,

pp 152–161

Ou Z & Chen B (2007) Stone arch bridges in Fujian, China. Proceedings of the

Arch 07 – 5th International Conference on Arch Bridges, PB Lourenço,

D Oliveira & V Portela (Eds), 12–14 September 2007, Madeira, pp 267–274

Paredes JA, Galindo JA & Mora DF (2007) Brick arch bridges in the high cauca

region of Colombia (1718–1919). Proceedings of the Arch 07 – 5th

International Conference on Arch Bridges, PB Lourenço, D Oliveira &

V Portela (Eds), 12–14 September 2007, Madeira. pp 193–202

Pauser A (2002) Massivbrücken ganzheitlich betrachtet. Geschichte – Konstruktionen

– Herstellung – Gestaltung. Österreichische Zementindustrie. Wien

Pauser A (2003) Konstruktions- und Gestaltungskonzepte im Brückenbau.

Betonkalender 2004, Part 1, Ernst und Sohn Verlag für Architektur und

technische Wissenschaften, Berlin, pp 27–96

Petersen C (1990) Statik und Stabilität der Baukonstruktionen. ´2., durchgesehene

Auflage, Vieweg Verlag

Petryna Y (2004) Schädigung, Versagen und Zuverlässigkeit von Tragwerken des

Konstruktiven Ingenieurbaus. Habilitationsschrift, Ruhr Universität Bochum –

Bauingenieurwesen, Schriftenreihe des Instituts für Konstruktiven

Ingenieurbau, Heft 2–2004, Shaker Verlag

Piecha A (1999) Die Begründbarkeit ästhetischer Werturteile. Fachbereich:

Kultur- u. Geowissenschaften, Universität Osnabrück, Dissertation

Prüfingenieur (2004) Der Prüfingenieur, April 2004, Nachrichten, p 7

Purtak F (2004) Zur nichtlineare Berechnung von Bogenbrücken aus

Natursteinmauerwerk. 22. CAD-FEM Users Meeting 2004 – International

Congress on FEM Technology with ANSYS CFX & ICEM CFD Conference,

10.–12. November 2004, Dresden

Radomski W (1996) Application of external prestressing for strengthening of

bridge structures in Poland. Recent Advances in Bridge Engineering.

Evaluation, management and repair, JR Casas, FW Klaiber & AR Marí (Eds).

Proceedings of the US-Europe Workshop on Bridge Engineering, organized

by the Technical University of Catalonia and the Iowa State University,

Barcelona, 15.–17. July 1996, International Center for Numerical Methods in

Engineering CIMNE, pp 597–612

REHABCON (2000) REHABCON IPS 2000–00063: Annex B: UK procedures

for bridge inspections, assessment and strengthening,

http://www.cbi.se/rehabcon/Annex_B.pdf, 2006

Rücker W, Hille F & Rohrmann R (2006) Guideline for the Assessment of

Existing Structures. SAMCO Final Report 2006

Ruddock T (1979) Arch Bridges and Their Builders 1735–1835. Cambridge

University Press

SABM (2006) The Stone Arch Bridge at Minneapolis, Minnesota.

http://voteview.com/stonearch.htm

Schaechterle K (1937) Fortschritte im Gewölbebrückenbau, Die Strasse, p 368

64 1 Introduction

Kunststeinmauerwerk, Die Bautechnik, p 145

Schaechterle K, Hammacher R, Brunkow W, Neuffer W & Seitz H (1956)

Brückenbau. Hütte - Des Ingenieurs Taschenbruch – Bautechnik. 28. Auflage.

Verlag von Wilhelm Ernst & Sohn, Berlin

Scheffler H (1857) Theorie der Gewölbe, Futtermauern und eisernen Brücken

sowol zum wissenschaftlichen Studium, als ganz besonders für den

praktischen Gebrauch der Ingenieure. Braunschweig, Schulbuchhandlung

Schlaich J (2003) Brücken: Entwurf und Konstruktion. Betonkalender 2004, Teil

1, Ernst und Sohn Verlag für Architektur und technische Wissenschaften,

Berlin, pp 1–26

Schmitt V (2004) Eisenbahnbrücken kleiner und mittlerer Stützweite im

Stahlverbund. 14. Dresdner Brückenbausymposium. 9. März, pp 115–121

Schneider J (1996) Sicherheit und Zuverlässigkeit im Bauwesen – Grundwissen

für Ingenieure. Vdf Hochschulverlag an der ETH Zürich und B.G. Teubner,

Stuttgart

Schubert JA (1847/1848) Theorie der Construction steinerner Bogenbrücken.

Theil 1: Dresden/Leipzig (Arnold) 1847; Theil 2: Dresden/Leipzig (Arnold)

1848; Figurentafeln: Dresden/Leipzig (Arnold) 1848

Schueremans L, Van Rickstall F, Ignoul S, Brosens K, Van Balen K & Van

Gemert D (2003) Continuous Assessment of Historic Structures – A State of

the Art of applied Research and Practice in Belgium. KULeuven, Department

of Civil Engineering

Seiler B (2006) Eritrea: Mit Mallets durchs Gebirge.

http://www.farrail.com/seiten/touren/eritrea-2006.html

Senker JH (2007) Pennsylvania stone arch bridges: evaluation and assessment.

Proceedings of the Arch 07 – 5th International Conference on Arch Bridges,

PB Lourenço, D Oliveira & V Portela (Eds), 12–14 September 2007, Madeira,

pp 257–265

Shrestha KM & Chen B (2007) An overview on “Arch bridges in Nepal”.

Proceedings of the Arch 07 – 5th International Conference on Arch Bridges,

PB Lourenço, D Oliveira & V Portela (Eds), 12–14 September 2007, Madeira,

pp 331–339

SIA 269 (2007) Grundlage der Erhaltung von Tragwerken. Draft Swiss code,

March 2007

Smith F (2003) The Reconstruction of brick arch bridges on the Chesterfield

canal. Structural faults + repair 2003. Abstracts of the tenth International

Conference of “Extending the life of bridges Concrete + Composites

Buildings, Masonry + Civil Structures”. The Commonwealth Institute,

London, p 65

Staudek T (1999) On Birkhoff’s Aesthetic Measure of Vases. Faculty of

Informatics, Masaryk University, Brno, Czech Republic

Steinbrecher D (2006) Geschichte des Holzbaus am Beispiel des

Holzbrückenbaues in Amerika und Europa. BTU Cottbus.

http://www.tu-cottbus.de/Stahlbau/lehre/holzbau

Straub H (1992) Die Geschichte der Bauingenieurkunst. Ein Überblick von der

Antike bis in die Neuzeit. Birkhäuser Verlag Basel

References 65

Vorlesungsbegleiter Teil III: Wirtschaftsgeschichte, Wintersemester 2003/04

THC (2006) The Heritage Council: Bridge Strengthening Procedures, Artificial

Roosts and Roost Construction Within Bridges: Bridge Strengthening

Procedures, Ireland, http://www.heritagecouncil.ie/publications/bats/ch6.html

Troyano LF (2003) Bridge Engineering – A Global Perspective. Thomas Telford

Turer A (2008) The need, challenges, and opportunities for research and

application of bridge health monitoring, a Turkish experience. Bridge

Maintenance, Safety, Management, Health Monitoring and Informatics, H Koh

& DM Frangopol (Eds), Taylor & Francis Group, London, pp 3063–3068

Tveit P (2007) Visit to the Steinkjer network arch 44 years later. Proceedings of

the Arch 07 – 5th International Conference on Arch Bridges, PB Lourenço,

D Oliveira & V Portela (Eds), 12–14 September 2007, Madeira, pp 305–314

UIC (2005) International Union of Railways: Improving Assessment,

Optimization of Maintenance and Development of Database for Masonry

Arch Bridges, http://orisoft.pmmf.hu/masonry/

Ural A & Dogangün (2007) Arch bridges in East Black sea region of Turkey and

effects of infill material on a sample bridge. Proceedings of the Arch 07 – 5th

International Conference on Arch Bridges, PB Lourenço, D Oliveira &

V Portela (Eds), 12–14 September 2007, Madeira, pp 543–550

Van der Vlist AA, Bakker MM, Heijnen WJ, Roelofs HJJ & Oosterhoff J (1998)

Bruggen in Nederland 1800–1940. Bruggen van beton, steen en hout.

Nedelandse Bruggen Stichting – Uitgeverij Matrijs

Velflík AV (1921) About four medieval stone bridges in Bohemia founded in

1169–1357. Technický obzor 18, pp 91–96

Vockrodt HJ (2005) Instandsetzung historischer Bogenbrücken im Spannungsfeld

von Denkmalschutz und modernen infrastrukturellen Anforderungen. 15.

Dresdner Brückenbausymposium, 15. März 2005, Dresden, pp 221–241

Vockrodt HJ, Feistel D & Stubbe J (2003) Handbuch Instandsetzung von

Massivbrücken. Untersuchungsmethoden und Instandsetzungsverfahren.

Basel, Boston, Berlin: Birkhäuser Verlag

von Hänsel-Hohenhausen M (2005) Vom Antlitz in der Welt. Nachwort im Buch:

Meine Lieder werden leben, Pohl, I. Frankfurter Verlagsgruppe, Frankfurt/M.

von Wölfel W (1999) Kanäle, Brücken und Zisternen - Berichte aus den antiken

Kulturen Ägyptens, Mesopotamiens, Südwestarabiens, Persiens und Chinas,

Bautechnik Spezial, Verlag Ernst & Sohn, Berlin

Wallsgrove J (1995) The Aesthetics of Load bearing Masonry Arch Bridges. Arch

Bridges, C Melbourne (Ed), Proceedings of the 1st International Conference

on Arch Bridges, Bolton 1995, Thomas Telford, London, pp 3–9

Weber WK (1999) Die gewölbte Eisenbahnbrücke mit einer Öffnung.

Begriffserklärungen, analytische Fassung der Umrisslinien und ein erweitertes

Hybridverfahren zur Berechnung der oberen Schranke ihrer Grenztragfähigkeit,

validiert durch einen Großversuch. Dissertation, Lehrstuhl für Massivbau der

Technischen Universität München

Wiebeking C-F (1809) Beyträge zur Brückenbaukunde, worin auch die neue

Bauconstruction wohlfeiler und dauerhafter Bogenbrücken, nach welcher

mehrere große Brücken vom Verfasser angegeben und ausgeführt sind,

66 1 Introduction

betrachtet werden können. München, Selbstverlag

Woodward RJ, Kaschner R, Cremona C & Cullington D (1999) Review of current

procedures for assessing load carrying capacity – Status C, BRIME PL97-2220.

January 1999

Yan BF & Shao XD (2008) Application of China bridge management system in

Qinyuan city. Bridge Maintenance, Safety, Management, Health Monitoring

and Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis Group,

London, pp 2675–2682

Yang Y, Chen B & Gao J (2007) Timber arch bridges in China. Proceedings of the

Arch 07 – 5th International Conference on Arch Bridges, PB Lourenço,

D Oliveira & V Portela (Eds), 12–14 September 2007, Madeira, pp 171–178

Yeomans D (2006) The safety of historic structures. The Structural Engineer, 21st

March, pp 18–22

Yi-Sheng M (1978) Bridges in China, old and new. From the Ancient Chaochow

Bridge to the Modern Nanking Bridge over the Yangtze. Foreign Languages

Press, Peking

Žnidarič A & Moses F (1997) Structural safety of existing road bridges. 7th

International Conference on Structural Safety and Reliability (ICOSSAR 97),

Kyoto, pp 1843–1850

Zucker P (1921) Die Brücke – Typologie und Geschichte ihrer künstlerischen

Gestaltung. Ernst Wasmuth a. G., Berlin

2 Loads

2.1 Introduction

As mentioned in the first chapter, bridges are designed to enable the trans-

port of goods over the shortest distance compared to the original geo-

graphical shape. Therefore, the bridge structure is exposed to certain loads,

not only the loads passing over the bridge, but also other types of loading

that are related to the inherent properties of bridges. Such loads include

temperature caused by changing weather conditions, the sun or snow, wind

loads, or simply the dead load of the bridge.

Historical arch bridges, in contrast to newly constructed bridges, require

further considerations of loading. For example, a historical bridge may be

unable to bear a modern traffic load, but the road can be weight restricted

to enable its continued use. In this chapter, loads are mainly considered in

reference to the viewpoint of historical arch bridges.

In the 1880s, Benz, Daimler, and Maybach more or less developed the pet-

rol driven car in parallel. This invention is the basis for the most modern

used road vehicle. However, that does not mean that only heavy trucks

have been used since this time. On the 11th of July 1893, the “Blue Won-

der Bridge” was opened in the German city of Dresden. The bridge was

first tested by many street car wagons loaded with rocks and anchors, in-

cluding several road rollers powered by steam engines, a street sprayer

pulled by horses, and one military company.

Neglecting earlier road traffic, the introduction of the petrol driven cars

can be seen as the beginning of an extremely successful growth of traffic

in many countries worldwide. This development, as shown in Figs. 2-1 and

2-2, began in Germany approximately 100 years ago with a marginal num-

ber of motorcars, by the 1950s already more than 2.5 million cars used in

Germany and nowadays about 44 million cars are reported (KBR 2001). The

D. Proske, P. van Gelder, Safety of Historical Stone Arch Bridges, DOI 10.1007/978-3-540-77618-5_2,

© Springer-Verlag Berlin Heidelberg 2009

68 2 Loads

maximum load for cars amounts to 44 tonnes. Lorries with higher loads are

permitted by special agreement. The number of allowances for such heavy

cars has increased in the last few years in Germany exponentially (Naumann

2002). On highways in Germany and Austria, lorries currently travel with a

weight of up to 100 tonnes (Hannawald et al. 2003, Pircher et al. 2009).

KBA (2001)

2.2 Road Traffic Loads 69

During the planning and design of road bridges, the engineer has to

forecast road traffic for decades to come. To simplify this task, different

traffic load models are included in codes of practice. Load models attempt

to model traffic loads on bridges on the one hand precisely, and on the

other hand, with a limited amount of work for the engineer. Good models

fulfill both of these requirements at the same time. The last requirement is

often questioned especially when Eurocode 1 traffic load models with

many different load combinations are considered. Independent from this

criticism, the intensive scientific work behind the models is appreciated.

Here only the exemplarily works by Merzenich and Sedlacek (1995) are

mentioned.

The road traffic model of the German DIN-reports 101 is heavily based

on the Eurocode 1 model, especially ENV 1991-3. This road traffic model

is valid for bridges with a maximum overall span and a maximum width of

42 m. A dynamic load factor is already considered in the characteristic

loads, if necessary. In contrast to the ENV 1991-3, the German DIN-report

101 considers only three road traffic load models: load model 1 with twin

axle and uniform distributed loads, load model 2 with a single axle for

short structural elements, and load model 4 to describe a human scrum. A

further load model in the ENV 1991-3 that considers special lorries has not

been considered in the German DIN-report 101.

Table 2-1 gives characteristic traffic load values for load model 1 and

Table 2-2 shows the distribution of the loads on a roadway according to

the DIN-report 101 and the historical DIN 1072.

Table 2-1. Characteristic loads for load model 1 according to the DIN-report 101

Load position Twin axle Uniform distributed load

Axle load Reduction factor qik Reduction factor

Qik × αQk

Qik in kN α Qk in kN/m2 αqk

Lane 1 300 0.8 240 9.0 1.0

Lane 2 200 0.8 160 2.5 1.0

Lane 3 100 0.0 0 2.5 1.0

Further lands 0 - 0 2.5 1.0

Remaining area 0 - 0 2.5 1.0

70 2 Loads

Table 2-2. Characteristic loads for load model 1 according to the DIN-report 101

and DIN 1072

DIN-report 101 DIN 1072

Characteristic loads for load model 1 in Characteristic loads for load model 1 in

the area of the twin axle the area of the SLW (Heavy Load

Vehicle)

Characteristic loads for load model 1 Characteristic loads for load model 1

outside the area of the twin axle outside the area of the SLW (Heavy

Load Vehicle)

Vehicle)

2.2 Road Traffic Loads 71

The road traffic model of the DIN 1072 includes two bridge standard

classes: the bridge class 60/30, which is used for new motorways, high-

ways, city roads (most roads); and the bridge class 30/30, which is used for

secondary roads. The bridge class 60/30 includes a main lane, a secondary

lane, and remaining areas, as does the load model 1. The main lane is ex-

2

posed to a uniformly distributed load of 5 kN/m and six single loads of the

SLW 60 (Heavy Load Vehicle). Furthermore, in the DIN 1072, the loads

are dependant on the span and the coverage increases by a dynamic load

factor. The secondary lane is exposed to a uniformly distributed load of

2

3 kN/m and six single axes of the SLW 30. No dynamic load factor is ap-

plied to the secondary lane and to the remaining areas. Also, the remaining

2

areas are exposed to a uniformly distributed load of 3 kN/m . In contrast to

the Eurocode or DIN-report 101 load model, the uniformly distributed

loads continue in the area of the SLW. Table 2-2 shows the load patterns

according to the DIN-report 101 and the DIN 1072.

Besides the two standard bridge classes, the DIN 1072 has introduced

bridge classes (BK 16/16, BK 12/12, BK 9/9, BK 6/6, BK 3/3) for check-

ing or recalibration (Table 2-3). Further historical load models for standard

20 and 8 tonnes lorries can be found in Leliavsky (1982).

Bridge class 16/16 12/12 9/9 6/6 3/3

Overall load in kN for the lorry 160.00 120.00 90.00 60.00 30.00

Front wheels Wheel load in kN 30.00 20.00 15.00 10.00 5.00

Contact width in m 0.26 0.20 0.18 0.14 0.14

Back wheels Wheel load in kN 50.00 40.00 30.00 20.00 10.00

Contact width in m 0.40 0.30 0.26 0.20 0.20

Single axle Wheel load in kN 110.00 110.00 90.00 60.00 30.00

Contact width in m 0.40 0.40 0.30 0.26 0.20

Uniform distributed load p1 in kN/m2 5.00 4.00 4.00 4.00 3.00

Uniform distributed load p2 in kN/m2 3.00 3.00 3.00 2.00 2.00

72 2 Loads

The DIN 1072 offers a wide range of different characteristic road traffic

loads and is therefore permitted a fine gradation for the usage of historical

bridges. This gradation cannot be found either in the Eurocode or in the

German DIN-report 101. If the codes of practice no longer offer special

load patterns for weight-restricted historical bridges, it would be helpful to

develop different characteristic road traffic loads for such weight-restricted

historical bridges.

There are many different theoretical scientific works about road traffic

models that can be used as a basis for such a development. Some works in

the German-speaking area were carried out by König and Gerhardt (1985),

Spaethe (1977), Schütz (1991), Krämer and Pohl (1984), Pohl (1993),

Puche and Gerhardt (1986), Bogath (1997), Bogath and Bergmeister (1999),

Ablinger (1996), Crespo-Minguillón and Casas (1997), and O’Connor and

O’Brien (2005).

Besides the Eurocode load model, different road models are used in

other countries, for example the HA-KEL and HB load model in Great

Britain, the HB-17 load model in the United States, or the load models T

44 and L 44 in Australia. The great diversity of load models is partially

caused by the high number of influencing parameters for traffic load pre-

diction models.

The development of traffic load models is, of course, strongly related to

the essential properties of road traffic. Road traffic is and will be (for an

indeterminate time) the most important means of traffic: it offers high

speed, high usability by the masses, and all uses omnipresent. On roads,

every non rail-tied vehicle can reach every road-realized goal at all times,

and the number of road-developed goals is immense compared to all other

means of traffic. This advantage of roads causes major drawbacks for the

road traffic models, since the numbers of influencing parameters is extre-

mely high. To develop models that can be used by engineers under prac-

tical conditions, the number of input parameters has to be strongly restricted.

Table 2-4 shows some load-influencing factors classified in four groups.

Table 2-4. Load-influencing factors for the estimation of the characteristic road

traffic loads according to Schütz (1991)

Traffic intensity Traffic flow Vehicle group Single vehicle

Average daily traffic in- Vehicle distance Frequency of Number of axles

tensity Lane distribution single vehicle Axle load

Average daily heavy Velocity types Axle distance

traffic intensity Vibration properties

Maximum hourly traffic

intensity

2.2 Road Traffic Loads 73

structural conditions have to be considered. Such parameters are the static

system of a bridge or the quality of the roadway. Such a quality of the

roadway can be considered in terms of local, regular, and irregular bumpi-

ness. A classification of the quality of the roadway in terms of the road

class is given in Table 2-5. A more detailed work about the road roughness

can be found in Bogsjö (2007).

Table 2-5. Roadway quality based on road classes (Merzenich and Sedlacek 1995)

Road class Roadway quality

Highway Excellent

Federal highway Good until very good

State road Good

Country road Average

for these parameters have to be found. These values are usually identified

by traffic measurement. However, although many measurement stations on

highways exist, the number on country roads associated with historical

arch bridges is rather limited. In Germany in 1991, about 300 measurement

stations were placed on highways and federal highways, but only 15 meas-

urement stations could be found on country roads (Loos 2005).

Also, on the European level, the major amount of traffic measurements

are carried out on highways, especially regarding the development of the

international European traffic load model, which focuses heavily on high

lorry traffic density measurements of long-distance traffic. Based on these

measurements, classes of lorries were identified. Merzenich and Sedlacek

(1995) have proposed four different lorry classes (Table 2-6).

Table 2-6. Suggestion for lorry classes according to Merzenich and Sedlacek (1995)

Class Description of lorry Representative lorry

Class 1 Lorry with two axes Two-axle vehicle

two axes

Class 3 Semi-trailer track Two axle track with

three axle semi-

trailer

Class 4 Tractive units Three-axle vehicle

with two axle with

trailer

74 2 Loads

is observed as shown in Fig. 2-3 (Geißler 1995, Quan 2004). This distribu-

tion consists of a distribution for the weights of both the unloaded and the

loaded lorries. Pohl (1993), for example, has considered in his model the

distributions of the weight of the lorry without load and the distribution of

the load itself, whereas other authors have considered axle loads as ran-

domly distributed variables (Fig. 2-4). An overview of these different

models can be found in Geißler (1995).

Fig. 2-3. General bimodal random distribution of the overall vehicle weight

(Geißler 1995, Quan 2004)

Fig. 2-4. General bimodal random distribution of the axle load (Geißler 1995)

neglects models for weight-restricted bridges, an extension of the Euro-

code road traffic model is recommended. However, the general procedure

outlined in the Eurocode for the development of traffic models will be

used. The goal of the following research is the development of α-factors

that can be applied to the Eurocode traffic model 1 to permit the re-

computation of load-restricted bridges.

To develop such a model, as previously suggested, requires measure-

ment data. Such traffic measurements were carried out on the Dresden

2.2 Road Traffic Loads 75

weight-restricted (15 tonnes) Blue Wonder Bridge (Fig. 2-5). Software was

used to process the axle weight measurement, including identification of ve-

hicle types. Such identification is mainly based on axle distances, and also

on the correlation between axle loads, as suggested by Cooper (2002). Re-

sults of the measurements of heavy weight vehicles are shown in Fig. 2-6.

Fig. 2-5. Blue Wonder Bridge in Dresden. Although the bridge is not a historical

arch bridge, it is a historical weight-restricted bridge and therefore it is assumed

that the traffic properties are comparable to historical arch bridges

76 2 Loads

Fig. 2-6. Relative frequency of measured overall vehicle weight and adjusted

multimodal normal distribution of heavy weight vehicles in October 2001 at the

Blue Wonder Bridge in Dresden

The factors that will be developed should deliver traffic models that are

comparable to the recalibration classes of the DIN 1072. The number in-

side the class gives the weight restriction for the lane, such as bridge class

30/30, 16/16 and 12/12.

Besides the input data obtained from measurement, a Monte Carlo

simulation is required. The input for the simulation is the decomposition

of the heavy vehicle-measured data classified into four lorry types (stan-

dard lorry or truck, truck with trailer, semi-trailer, and busses). Seeing

that the measurement data from the Blue Wonder does not include all

relevant information, further data were taken from Merzenich and Sed-

lacek (1995). The approximation of the axle load distribution was done

by bimodal distribution.

In general, the α-factors will be determined in the following steps:

1. Simulation of the vehicle type based on the traffic contribution of the

Blue Wonder Bridge data.

2. Simulation of the overall vehicle weight based on the available Blue

Wonder Bridge data on vehicle weights.

3. Simulation of the axle load contributions to the overall vehicle weight

based on the work by Merzenich and Sedlacek (1995).

2.2 Road Traffic Loads 77

4. Computation of the axle loads based on the overall vehicle weight and

the axle-load contribution.

5. Simulation of the axle distances based on Merzenich and Sedlacek

(1995).

6. Computation of the maximum bending moment for a single-span beam.

There are some simplifications involved in the simulation process. For

example, five-axle vehicles are simplified so that the fifth axle has the

same weight as the fourth axle. Also, the decisive load pattern was identi-

fied by iteration. The simulation itself was repeated 5,000 times for one

length of a single-span beam. Of course, the length of the single-span

beam was varied. For the maximum bending moment of a single-span

beam with a certain length, the simulation yielded to a frequency distribu-

tion. A log-normal distribution has been applied parallel to the approxima-

tion of the frequency data by a normal distribution. The characteristic traf-

fic load value is assumed as a value with a 1,000-year return period of this

distribution (Merzenich and Sedlacek 1995).

After the computation of the maximum bending moment by the simula-

tion process, the α-factor was computed by the required adaptation of the

standard traffic model 1 according to the Eurocode 1 and the DIN-report

101. The standard traffic model 1, including the α-factor, should give com-

parable results in terms of moments as the simulated computation.

Together with flowing traffic conditions, traffic jam conditions must

also be considered. Traffic jam conditions are mainly relevant for long-

span conditions and have to be considered in the computation of the

factors.

The described procedure was applied for the bridge class 16/16. How-

ever, for the bridge classes 12/12 and 30/30, the procedure had to be slightly

changed, since measurements were only based on the bridge class 16/16.

For the bridge class 12/12, the mean value of the measurement data

from the Blue Wonder Bridge was multiplied with 12/16 = 0.75. The stan-

dard deviation and the contribution of the different vehicle types to the

traffic were kept constant. For the adaptation of the bridge class 30/30, this

procedure was again extended. Based on measurements from the Blue

Wonder and Auxerre-traffic (Fig. 2-7), a new overall traffic weight distri-

bution was constructed by changing mean values, standard deviation, and

contribution of different vehicle types to the traffic from the Blue

Wonder Bridge traffic (Fig. 2-8). The simulation was then repeated.

To prove the suggested approach, the α-factor of 1.0 for unified traffic

load in the traffic model 1 in the Eurocode was verified for the Auxerre-

traffic (Fig. 2-9). For the axle load, a value of 0.9 was found. However,

this slight difference can be interpreted as an additional safety element.

The computed α-factors are summarized in Table 2-7.

78 2 Loads

Table 2-7. Reduction factor α for different bridge classes for the new

Bridge class Roadway Lane 1 Lane 2

quality

αQ1 αq1 αQ2 αq2

3/3* Average 0.1 0.22

6/6* Average 0.2 0.24

9/9* Average 0.25 0.26

12/12 Good 0.30 0.28 0.20 1.00

Average 0.30 0.30 0.25 1.00

16/16 Good 0.35 0.30 0.35 1.00

Average 0.35 0.40 0.45 1.00

30/30 Good 0.55 0.70 0.50 1.00

Average 0.60 0.70 0.80 1.00

Simulation “Auxerre-traffic” Good 1.00 0.90 1.00 1.00

Load model 1 DIN-report 101 0.80 1.00 0.80 1.00

*First drafts

Fig. 2-7. Comparison of overall vehicle weights measured at the Blue Wonder

Bridge in Dresden and at the Auxerre-traffic in France. The latter was mainly used

for the development of the Eurocode traffic model 1

2.2 Road Traffic Loads 79

Fig. 2-8. Development of a synthetic traffic distribution for the bridge class 30/30

based on measurements from the Blue Wonder in Dresden and from the Auxerre-

traffic in France based on Loos (2005)

Fig. 2-9. Maximum bending moments caused by characteristic traffic loads, in-

cluding dynamic load factor according to Loos (2005)

80 2 Loads

To provide further control of the suggested factors, the results have also

been compared with the road traffic model by Pohl (1993). Pohl distin-

guishes between long-distance, average-distance and short-distance traffic.

Long-distance traffic represents more or less heavy vehicle traffic on

German highways. Average-distance traffic can be found on federal high-

ways and on country roads, and short-distance traffic can be found on

weight-restricted routes. Therefore, the simulation procedure using the

model of Pohl is slightly different than the simulation procedure explained

above. Furthermore, in our simulation, the short-distance traffic is sepa-

rated into two types (types 1 to 4 or types 1 and 2 according to Table 2-6).

Figure 2-9 shows the characteristic maximum bending moment of a

single-span beam for different spans and different load models. It permits

a direct comparison of the load model by Pohl, the Eurocode model, and

the suggested variation of the Eurocode model.

The factors given in Table 2-7 depend on the roadway quality. Usually

this property is not given in codes, however here it is assumed that for

country roads lower roadway quality can be found that has significant im-

pacts on the chosen α-factor due to the dynamic properties. Here, the

model from Merzenich and Sedlacek (1995) has been applied. In general,

the different roadway qualities do not have such a strong effect on the

main lane, however the second lane is influenced more by this property.

It should be stated here that the factors found are a much more appropri-

ate method for the recomputation of historical bridges than the sometimes

found simple diminishing factors of the Eurocode load model 1 of 0.9, 0.8,

0.7 and so on, as has been suggested in Vockrodt (2005). Applying these

factors of 0.9, 0.8 or 0.7 to the Eurocode model 1 violates general assump-

tions in the statistical properties of model 1.

Further adaptations of the Eurocode model 1 are known, or different

models have been proposed besides the presented schematic. Such an addi-

tional adaptation has been presented by Novák et al. (2007).

A different proposal for the consideration of local traffic conditions for

local traffic loads has been offered by Bailey and Hirt (1996). This method

shows the corresponding stochastic basis much stronger than the Eurocode

model 1. The adaptation of traffic load to local traffic according to Bailey

and Hirt (1996) considers

• Maximum overall lorry weights in qmax in kN/m

• Mean overall lorry weights μQ in kN/m

• Standard deviation of the overall lorry weights σQ in kN/m

• Traffic volume N

2.2 Road Traffic Loads 81

• Proportion of the freely flowing traffic F on traffic volume

Based on these data, six coefficients are computed:

q (2-1)

c1 = max ⋅ 0, 2 + 0,8 (40 ≤ qmax in kN/m ≤ 80)

73

1 (2-2)

c2 = (6 ≤ μQ in kN/m ≤ 20)

μQ

⋅ 0,65 + 0,35

14,5

1 (2-3)

c3 = (2 ≤ σ Q in kN/m ≤ 8)

σQ

⋅ 0,6 + 0, 4

6,0

1 (2-4)

c4 = (0,1 ≤ HV ≤ 0, 4)

HV

⋅ 0,7 + 0,3

0, 25

1 (2-5)

c5 = (105 ≤ N ≤ 109 )

log( N ) ⋅ 0,08 + 0,33

F (2-6)

c6 = ⋅ 0, 2 + 0,8 (40 ≤ F ≤ 100)

94

6 ⋅ c1 ⋅ c2 ⋅ c3 ⋅ c4 ⋅ c5 ⋅ c6 (2-7)

αQ =

( c1 + c2 + c3 + c4 + c5 + c6 )

Q (2-8)

Q= d

αQ

The example in Table 2-8 is taken from Bailey and Hirt (1996). The co-

efficients are then c1 = 0.99; c2 = 1.02; c3 = 1.0; c4 = 1.72; c5 = 1.09; c6 =

1.01, and finally αQ = 1.68.

82 2 Loads

Local traffic properties Value

Maximum overall weight of lorries in qmax in kN/m 70

Mean overall weight of lorries μQ in kN/m 14

Standard deviation of the overall weight of lorries σQ 6

(kN/m)

Traffic volume N 20 × 106 in 10 years

Proportion of heavy truck traffic HV on the traffic 0.05 rounded up to 0.1

Proportion of the freely flowing traffic F on the traffic 1% standing

volume 2% with 40 km/h

500 vehicles per hour

The major problem of this schematic is probably the capture of the data.

Many publications concerned with traffic loads can be found. Exemplary

publications are by Casas and Crespo-Minguillon (1996), Hannawald et al.

(2003), COST-345 (2004), and Allaix et al. (2007). Further traffic load

models can be found in COST-345 (2004), Vrouwenvelder and Waarts

(1993), and Prat (2001).

He simply put a steam engine on wheels. At first, railways were used for

goods traffic, but have also been used since 1829 for passenger traffic,

when Robert Stephenson won the race. In the following 100 years, rail-

roads experienced incredible growth (Table 2-9).

Year Railroad length in Germany in km

1840 549

1850 6044

1870 19575

1910 61148

construction. Therefore, between 1845 and 1890 especially, many railway

bridges were constructed as arch or vault bridges. Later, most superstruc-

tures were built from steel.

Because of the rapidly growing demand for railway passenger and

goods transport, locomotives were permanently improved. This improve-

ment was related to an increased overall weight. For example, from 1835

2.3 Railroad Traffic Load 83

corresponds to an increase in uniform load from 2.5 to 13.67 tonnes per

metre (Beyer 2001). As an example, the growth of railway loads in Spain

is shown in Table 2-10.

Year Uniform load in kg/m

1877 4,900

1902 6,230

1925 10,780

1956 13,400

1975 12,000

Germany (1845–1876), the railway load design was a train with real axle

loads. A locomotive with an overall length of 7.0 m and wheel distances of

2.4 and 1.4 m was used. The wheel load was 5.0 and 4.5 tonnes, respec-

tively. The maximum speed was 40 km/h. Only in 1877, the Wurttemberg

railway company introduced a railway load pattern.

The use of the railway load pattern UIC 71 for new bridges was devel-

oped deterministically based on the sum of load patterns of real railway

loads concurrently in Germany and many other European countries. It was

introduced in 1971. The introduction of a new load pattern for railways

called IM 2000 has so far failed. In addition to the general railway loads

for heavy railways, the so-called, heavy load patterns SW/0 and SW/2 are

used. For consideration of local conditions, the UIC 71 can be adapted by

factor α. The factor can reach from 0.75 to 1.33 (see also Fig. 2-10). This

would permit an adaptation of the load for special conditions on historical

bridges.

If this adaptation is insufficient, then there exists a further opportunity

by using special static standard railway load models. These special railway

models are related to lower railway track classifications. Figure 2-11 in-

cludes examples of such lower classifications and the standard load pat-

terns. However, the application of such load patterns requires a strong col-

laboration with the railway company (in Germany the DB AG, O’Connor

et al. 2008).

84 2 Loads

Fig. 2-10. Observed moments on bridges in comparison with the assumed moments

based on the UIC loads (Lieberwirth 2004)

2.3 Railroad Traffic Load 85

Fig. 2-11. Railway load patterns for the UIC 71 and the recomputation railway

patterns C3 and D4 according to the DB AG. Furthermore, the possible future load

patterns IM-2000 (Weber 1999) and the historical load pattern G1 are shown

mainly driven by the first steam locomotives, the heavy goods railway

transport has significantly changed. Locomotives are much stronger com-

pared to the old days, which yields not only to higher starting tensile forces

but also to heavier railway wagons. Nowadays, the axle loads of railway

wagons exceed the axle loads of locomotives and show much higher devia-

tions caused, for example, by overload or humidity of the bulk freights.

An international load border system for goods railway wagons gives the

maximum load class, up to which wagons on certain railway routes can be

loaded (Tables 2-11 and 2-12).

Table 2-11. Load border system for goods railway wagons (t = tonnes)

A/B1 B2/C2 C3/C4 D2 D3 D4

38.0 t 56.0 t 65.0 t 56.0 t 67.0 t 77.0 t

Maximum speed 120 km/h

Table 2-12. Interpretation of the load border system in Germany (DB AG)

Classification Wheel set load

Uniform vehicle load A B C D

in tonne/m 16 tonnes 18 tonnes 20 tonnes 22.5 tonnes

1 5.0 A B1

2 6.4 B2 C2 D2

3 7.2 C3 D3

4 8.0 C4 D4

86 2 Loads

The load border system influences the railway traffic on a certain route

in different ways—for example, the speed can be limited. This influences

the dynamic load factor and therefore decreases the load on structures on

this route. Additionally, railway goods trains can only be loaded to a cer-

tain value. Usually, the wheel set load is declared on the goods wagon and

the wagons are only then loaded up to the permitted value corresponding

to the load border system.

Since the route classification often depends on the weakest link, which

is usually a bridge, it becomes clear that great interest from railway com-

panies exists to utilize all load-bearing capabilities of historical arch

bridges in order to reach an acceptable route class. If the historical arch

bridge remains an impairment for the route classification, then sooner or

later the bridge will be replaced. However, this problem provides an op-

portunity to apply modern concepts for both the load description and the

structural resistance description.

Background information about the European railway loads can be found

in some ERRI reports (1993, 1994a, 1994b and 1998), UIC-report 702

(1974), Tobias et al. (1996), or Lieberwirth (2004).

Initial drive forces from railway trains and road trucks can introduce high

main spin direction parallel horizontal forces. However, this depends on

the structural components of the arch superstructure. In Germany, single-

span arch bridges with sufficient coverage have rather low average span

distances, continuous rail initial drive forces, and breaking forces. This is

based on the Ril 805 (1999) and Ril 804 (2003) for the recomputation of

historical railway bridges.

Weber (1999) assumes that by using continuous rails, the initial drive

forces and the breaking forces are directly transferred to the connected

railway. Therefore, these forces are not transferred by the arch bridges and

are not measurable in the arch itself. A historical draft of the Eurocode 1,

Sect. 3.4, had assumed that initial drive forces and breaking forces do not

have to be considered for historical vault and arch bridges.

Ril 804 considers historical bridges by reducing the maximum initial

drive force of 1 MN by a factor ξ.

Fx = 33,3 ⋅ l ⋅ ξ (2-9)

2.5 Breaking Forces 87

For long arch lines, special considerations are required since the loading

length can easily exceed the above-mentioned length. Here, for example,

the type of construction of the arch and the piers has to be considered. It

should be noted that even when a lower load pattern than UIC 71 is used

for the estimation of vertical loads, the initial drive force must be consid-

ered. This is understandable if one considers that the strongest locomotives

in Germany are used in the train route class D4. Currently, the strongest

locomotive of the DB AG (Germany) is the modernized BR 241

(Worowschilowgrad). This locomotive is mainly used for heavy goods

trains and reaches a maximum pull force of 450 kN over 20.82 m.

based on the same function as the initial drive force:

Fx = f x , Br ⋅ l ⋅ ξ in kN. (2-10)

For every track, fx,Br = 20 kN/m has to be used for passenger and good

trains. However, for heavy goods transport, an increased value of 35 kN/m

has to be applied since the heavier wagons can cause greater breaking

forces.

For tall and slender piers, a load combination of relatively low vertical

traffic load (low route class) with rather high initial drive or breaking

forces and side wind is often relevant for design. To fulfill the stability re-

quirements for long bridges, different pier types are often used: these are

called group piers, which are thicker than the normal piers and have the

function to transfer high horizontal forces to the foundations. The result is

that only one group pier has to take the horizontal forces between the

group piers. The coupling of the different bridge parts needs to then be

proven independently. For the coupling, the backfill concrete on the arches

is often considered. Post-applied precast concrete way elements are usually

not of considerable strength for the horizontal force transfer due to the

joints.

Breaking forces for road bridges were originally computed based on the

DIN 1072 (1985) and supplementary sheets. Values range from approxi-

mately 10 to 900 kN. Currently, the Eurocode 1 or the DIN-report is the

basis for the estimation of the breaking forces. Figure 2-12 shows breaking

force values subject to the bridge length and different vehicle types.

88 2 Loads

Fig. 2-12. Breaking forces on bridges subject to the bridge length and the type of

design vehicle according to Merzenich and Sedlacek (1995)

forces into the foundation requires models that consider the stiffness of the

single piers. Further work on this topic can be found in Merzenich and

Sedlacek (1995), Pfohl (1983), and Weihermüller and Knöppler (1980).

Arch bridges are exposed not only to horizontal forces parallel to the main

spin of the bridge but also to horizontal forces rectangular to the main spin

of the bridge. Such forces are impact forces or wind forces. Wind forces

indeed act on historical arch bridges, however in most cases the wind

forces are not relevant due to the high dead load of arch bridges. As men-

tioned before, sometimes wind forces have to be considered in combina-

tion with other horizontal forces, mainly for tall piers.

Codes of practice for wind forces on bridges are again the Ril 804 and

Ril 805 for railway bridges in Germany, and the DIN-report 101. It should

be mentioned that often wind forces with and without traffic on bridges

have to be considered separately.

There are many background documents about wind force evaluation in-

cluding Lieberwirth (2003).

2.9 Temperature Loading 89

caused by technical and natural means. Technical means are mainly differ-

ent means of transport, such as impacts from ships, cars, railways, or air-

planes and helicopters. Natural processes are mainly gravitational mass

movements such as avalanches, debris flows, flash floods, landslides, or

ice pressure.

Many historical bridges have been destroyed by flash floods or other

impacts. Such forces can easily reach several Mega Newton. Background

for the computation of impact forces by means of transport can be found in

Proske (2003), and for natural processes in Proske and Hübl (2007) and

Proske et al. (2008).

2.8 Settlements

The static integrity of the arch bridge is strongly related to the low de-

formation of the foundation and the abutment. This is the reason why

arch bridges are usually not built in regions with soft ground, such as

gravel or sand. Therefore, settlements of the foundation of arch bridges

can cause substantial damage to the bridge. However, for historical arch

bridges, settlements usually should not increase further and if the bridge

has survived a number of decades or centuries, settlement should no

longer be of concern.

The consequences of settlements are discussed, for example, in Jagfeld

(1998, 2000) and Ochsendorf (2002). Ochsendorf gives an all-inclusive

number for the decrease of the ultimate load-bearing behaviour of arch

bridges of 15%. Furthermore, settlements can change the location of the

hinges inside the arch and can therefore completely change the failure

mechanisms. Goldschneider (2003) discusses the combinations of ground

and historical foundations.

puted due to a number of different opinions.

Based on the UIC-report, a temperature gradient over the vault cross

section has to be considered. The constant temperature change has to be

90 2 Loads

former Eurocode draft.

For temperature loading, Pietsch (1961) recommends either a computa-

tion with consideration of the plasticity of masonry or an artificial decrease

of the stiffness E0 of the masonry, depending on the temperature change.

He proposes for the

daily temperature change, Et = 1,0 ⋅ E0 , (2-11)

This yields to rather low internal forces inside the arch due to tempera-

ture loading. The factor of 0.2 fits very well with a publication by Thürmer

(1995). He calculated a loss of stress caused by temperature inside a his-

torical arch bridge by a nonlinear computation of 70% compared to the

linear-elastic computation. This would represent a factor of 0.3.

Bothe et al. (2004) presents the computation of a historical arch bridge.

He calculates that a linear-elastic computation without temperature loading

results in a maximum live load of 56 kN/m and consideration of tempera-

ture loading results in a 14 kN/m maximum live load. However, if a

nonlinear computation based on maximum strain and overall stability is

carried out, the maximum live load becomes independent from the tem-

perature loading. If a certain characteristic concrete strength is assumed for

the resistance, then the maximum live load without temperature loading is

102 kN/m, compared to 100 kN/m with temperature loading. Figure 2-13

shows the formulae for the computation of the characteristic concrete

strength, including further additional formulas. Besides the three nonlinear

computations with and without temperature, the linear-elastic computation

results are also included in the figure. The diagram shows clearly that the

consideration of temperature loadings on historical arch bridges does not

result in a significant decrease of the ultimate load-bearing capacity, if

nonlinear computations are carried out.

2.9 Temperature Loading 91

Nonlinear

computation

Fig. 2-13. Computation of maximum permitted live loads on a historical arch con-

sidering four load cases. First, a linear-elastic computation is carried out consider-

ing temperature and no temperature. Furthermore, three nonlinear computations

with and without temperature are shown. These computations differ in the as-

sumption of the concrete compression strength fc as shown. As we can see, α is the

long-term loading factor (0.85), fck is the characteristic concrete compression

strength, and γR is the partial safety factor (1.8)

This statement fits very well with the regulation of the historical DIN

1072 (1952), whereon masonry historical arch bridges with backfill, func-

tional arch without hinges, a maximum span of 20.0 m, and with a certain

pier ratio temperature loading had not to be considered.

In contrast to this, there are some known works where temperature loads

caused significant effects on historical arch bridges. Patzschke (1996) de-

scribes an example where temperatures in the summer of 1992 caused the

appearance of a three-hinge arch on a railway multi span brick arch bridge.

Since the development of hinges in the arch caused major deformations

and limited the use of this railway route, structural hinges were installed to

limit the deformations. Schlegel et al. (2003) also mention that temperature

caused major deformations on the Göltzschtal Bridge in Saxony.

Further investigations into the influence of temperature loadings on

spandrel walls and parapets are given by Robinson et al. (1998). They use

ABACUS to simulate climatic effects on arch bridges. The temperatures

considered range from 0° to 31°C. Measurements were also considered in

this study.

92 2 Loads

since no traffic can pass the bridge under high snow loads. However, the

traffic weight is usually much higher than the snow load and therefore

dominates design. Further information about snow load can be found in

Lieberwirth (2003), JCSS (2004), Soukhov (1998), and Mehlhorn (1997).

All loads discussed so far have been live loads. They change direction and

presence. However, the most important load on arch bridges is the dead

load or self-weight. In contrast to many other structures, the dead load only

allows the function of the arch. If the self-weight is missing, then the sta-

bility of the arch itself is not given.

Therefore, the correct computation of the dead load is of utmost impor-

tance. For these computations, the density and volume of certain structural

elements and materials are required. The volume is mainly computed on

historical or up-to-date documentation about the structure. Although the

computation of the volumes and weight is rather simple, one should care-

fully investigate documents or the structure about the assembly and the ge-

ometry. In some cases, drilling or other further investigation technologies

may be required to find cavities in the structure. Such cavities can be blast-

ing chambers or vaults inside the structure.

Once the geometry is known, either based on measurements or based on

codes of practice like the DIN 1055-1, the density of the material can be

chosen. As an example, for the bridge shown in Fig. 2-14, the following

densities were used:

3

Density Porphyr gP =28 kN/m (DIN 1055-1, 2002, Tabelle 6, Zeile 12),

3

Density Sandstone gS =27 kN/m (DIN 1055-1, 2002, Tabelle 6, Zeile 15),

3

Density Concrete gB =24 kN/m (DIN 1055-1, 2002, Tabelle 1, Zeile 21).

2.11 Dead Load 93

Fig. 2-14. Example geometry draft of a bridge (no internal volumes are shown)

If the densities cannot be taken from the codes, mean values of meas-

urement data can be taken. The mean value as characteristic values for

densities and volumes correspond with Eurocode 1, if the coefficient of

variation is less than 0.05. If the deviation is higher, then 95% fractile val-

ues can be chosen. However, this is rather unusual, since slight deviations

are covered by a partial safety factor of 0.95 and 1.05.

Although partial safety factors are discussed later in Chapter 7, some

remarks should be given here, since the dead load on arch bridges is a very

specific case. This is caused by the fact that the dead load is increasing the

ultimate load-bearing capacity of arches up to a certain value. Over a cer-

tain value, the dead load becomes like the live load, and decreases the ul-

timate load-bearing behaviour.

Whereas in the Ril 805 for the German railway bridges the partial safety

factor for the dead load is γG = 1.20, in the Eurocode 1 or in the German

DIN 1055-100, γG = 1.35. Works by Wiese et al. (2005) have shown that

this value is too high and considers not only uncertainties from the dead

load but also from the modelling. A former draft of the Eurocode gives

some special considerations to historical arch bridges:

• The partial safety factor for the self-weight of the arch can be decreased

to 1.1 for unfavourable effects and 1.0 for favourable effects. Herrbruch

et al. (2005) have suggested 1.35 and 0.9, respectively, according to the

current codes.

• The partial safety factor for the self-weight of further elements of the

superstructure (like coverings) can be chosen as 1.2 for unfavourable

and 0.9 for favourable effects.

94 2 Loads

References

und stochastische Modelle, Diplomarbeit, Institut für konstruktiven

Ingenieurbau, Universität für Bodenkultur

Allaix DL, Vrouwenvelder ACWM & Courage, WMG (2007) Traffic load effects

in prestressed concrete bridges. 5th International Probabilistic Workshop,

L Taerwe & D Proske (Eds), Ghent, Belgium, pp 97–110

Bailey SF & Hirt AH (1996) Site specific models of traffic action effects for

bridge evaluation. Recent Advances in Bridge Engineering. Evaluation,

Management and Repair, JR Casas, FW Klaiber & AR Marí (Eds),

Proceedings of the US-Europe Workshop on Bridge Engineering, organized

by the Technical University of Catalonia and the Iowa State University,

Barcelona, 15–17 July 1996, International Center for Numerical Methods in

Engineering CIMNE, pp 405–425

Beyer P (2001) 150 Jahre Göltzschtal- und Elstertalbrücke im sächsischen

Vogtland 1851–2001. Vogtland Verlag Plauen

Bogath J & Bergmeister K (1999) Neues Lastmodell für Straßenbrücken.

Bauingenieur 6, pp 270–277

Bogath J (1997) Verkehrslastmodelle für Straßenbrücken. Dissertation, Institut für

konstruktiven Ingenieurbau, Universität für Bodenkultur

Bogsjö K (2007) Evaluation of stochastic models of parallel road tracks.

Probabilistic Engineering Mechanics, 22, No. 4, October 2007, pp 362–370

Bothe E, Henning J, Curbach M, Bösche T & Proske D (2004) Nichtlineare

Berechnung alter Bogenbrücken auf der Grundlage neuer Vorschriften. Beton-

und Stahlbetonbau 99, Heft 4, pp 289–294

Casas JR & Crespo-Minguillon C (1996) Serviceability and fatigue evaluation of

existing bridges. Recent Advances in Bridge Engineering. Evaluation,

Management and Repair, JR Casas, FW Klaiber & AR Marí (Eds), Pro-

ceedings of the US-Europe Workshop on Bridge Engineering, organized by

the Technical University of Catalonia and the Iowa State University,

Barcelona, 15–17 July 1996, International Center for Numerical Methods in

Engineering CIMNE, pp 441–464

Cooper D (2002) Traffic and moving loads on bridges. Dynamic Loading and

Design of Structures, AJ Kappos (Ed), Chapter 8, Spon Press, London,

pp 307–322

COST 345 (2004) European Commission Directorate General Transport and

Energy: Procedures Required for the Assessment of Highway Structures:

Numerical Techniques for Safety and Serviceability Assessment – Report of

Working Groups 4 and 5

Crespo-Minguillón C & Casas JR (1997) A comprehensive traffic load model for

bridge safety checking, Structural Safety, 19 (4), pp 339–359

DIN-report 101: Einwirkungen auf Brücken. Beuth Verlag, Berlin, 2001

ENV 1991-3 Eurocode 1: Grundlagen der Tragwerksplanung und Einwirkungen

auf Tragwerke, Teil 3: Verkehrslasten auf Brücken, 1995

ERRI-report (1993) D 192/RP 1: Theoretische Grundlagen zur Überprüfung des

bestehenden Lastbildes UIC 71, Utrecht

References 95

und künftigen Schienenverkehrs auf den internationalen Strecken mit denen

aus Lastbild UIC 71 auf deterministischer Grundlage, Utrecht

ERRI-report (1994b) D 192/RP 3: Lastbild für die Berechnung der Tragwerke der

internationalen Strecken, Utrecht

ERRI-report (1998) D 221/RP 1: Vertikale Eisenbahnverkehrslasten:

Kombinationsbeiwerte – Hintergrundinformationen zum Thema

“Eisenbahnverkehrslasten”, Utrecht

Geißler K (1995) Beitrag zur probabilistischen Berechnung der Restnutzungsdauer

stählerner Brücken. Dissertation, Heft 2 der Schriftenreiche des Institutes für

Tragwerke und Baustoffe an der Technischen Universität Dresden

Goldschneider M (2003) Baugrund und historische Gründungen. Untersuchen,

Bewerten und Instandsetzen. Empfehlungen für die Praxis, Fritz Wenzel und

Joachim Kleinmanns (Hrsg) “Erhalten historisch bedeutsamer Bauwerke”.

Karlsruhe

Hannawald F, Reintjes K-H & Graße W (2003) Messwertgestützte Beurteilung

des Gebrauchsverhaltens einer Stahlverbund-Autobahnbrücke. Stahlbau 72

(7), pp 507–516

Herrbruch J, Groß JP & Wapenhans W (2005) Wirklichkeitsnähere

Tragwerksanalyse von Gewölbebrücken. Mauerwerk 9, Heft 3, pp 89–92

Jagfeld M (1998) Tragverhalten gemauerter Wölbkonstruktionen bei großen

Auflagerverschiebungen – Vergleich verschiedener FE Modelle. 16. CAD-

FEM User’s Meeting, 7.–9. Oktober 1998 in Bad Neuenahr

Jagfeld M (2000) Tragverhalten und statische Berechnung gemauerter Gewölbe

bei großen Auflagerverschiebungen – Untersuchungen mit der Finit-Element-

Methode. Schriftenreihe des Lehrstuhls für Hochbaustatik und

Tragwerksplanung, Technische Universität München, Aachen

JCSS (2004) Joint Committee of Structural Safety (2004). Probabilistic

Modelcode. www.jcss.ethz.ch

KBA (2001) Kraftfahrzeugbundesamt: Festschrift 50. Jahre Kraftfahrt-Bundesamt

1951–2001. August 2001

König G & Gerhardt HC (1985) Verkehrslastmodell für Straßenbrücken.

Bauingenieur 60, pp 405–409

Krämer W & Pohl S (1984) Der Ermüdungsnachweis in dem Standard TGL

13460/01. Ausgabe 1984 – Grundlagen und Erläuterungen. Die Straße 24,

Heft 9, pp 257–263

Leliavsky S (1982) Arches and short span bridges. Design Textbook in Civil

Engineering: Volume VII, Chapman and Hall, London

Lieberwirth P (2003) Ein Beitrag zur Wind- und Schneelastmodellierung. 1.

Dresdner Probabilistik-Symposium – Sicherheit und Risiko im Bauwesen,

D Proske (Ed), Technische Universität Dresden, Dresden, pp 123–138

Lieberwirth P (2004) Ausgewählte Aspekte zur statistischen Beschreibung von

Bahnverkehrslasten. 2. Dresdner Probabilistik Symposium, Dresden

Loos S (2005) Entwicklung von Lastmodellen auf Grundlage des DIN-

Fachberichtes 101 für gewichtsbeschränkt beschilderte Straßenbrücken.

Diplomarbeit. Technische Universität Dresden, Institut für Massivbau

96 2 Loads

Achter Band. Propyläen Verlag Berlin – Frankfurt am Main

Mehlhorn G (1997) Der Ingenieurbau, Band 8 – Tragwerkszuverlässigkeit und

Einwirkungen, Verlag Ernst & Sohn, Berlin

Merzenich G & Sedlacek G (1995) Hintergrundbericht zum Eurocode 1 – Teil 3.2:

Verkehrslasten auf Straßenbrücken. Forschung Straßenbau und Straßenver-

kehrstechnik. Bundesministerium für Verkehr, Heft 711

Naumann (2002) Aktuelle Schwerpunkte im Brückenbau. 12. Dresdner

Brückenbausymposium. Lehrstuhl für Massivbau und Freunde des

Bauingenieurwesens e.V., pp 43–52

Novák B, Brosge B, Barthel K & Pfisterer W (2007) Anpassung des

Verkehrslastmodells des DIN FB-101 für kommunale Brücken, Beton- und

Stahlbetonbau 102, Heft 5, pp 271–279

O’Connor A & O’Brien EJ (2005) Traffic load modelling and factors influencing

the accuracy of predicted extremes. Canadian Journal of Civil Engineering 32,

pp 270–278

O’Connor A, Pedersen C, Enevoldsen I, Gustavsson L & Hammarbäck J (2008)

Probability Based Assessment of a Large Riveted Truss Railway Bridge.

Bridge Maintenance, Safety, Management, Health Monitoring and

Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis Group,

London, pp 1812–1819

Ochsendorf JA (2002) Collapse of Masonry Structures. University of Cambridge,

Dissertation, Cambridge

Patzschke F (1996) Der Wahrener Viadukt in Leipzig – Rekonstruktion einer

Gewölbebrücke der DB AG. 6. Dresdner Brückenbausymposium, Fakultät

Bauingenieurwesen, Technische Universität Dresden, pp 135–148

Pfohl H (1983) Reaktionskraft am Festpunkt von Brücken aus Bremslast und

Bewegungswiderständen der Lager. Bauingenieur 58, pp 453–457

Pietsch H (1961) Zur Frage der Beanspruchung aus Temperatur in massiven

Gewölbebrücken am Beispiel des voll eingespannten Bogens mit

Übermauerung. Dissertation Hochschule für Verkehrswesen, Dresden

Pircher M, Lechner B, Mariani O & Kammersberger A (2009) Schädigung einer

schlaff bewehrten Betonbrücke durch Verkehrsbelastung. Beton- und

Stahlbetonbau, 104(3), pp 154–163

Pohl S (1993) Definition von charakteristischen Werten für Straßenverkehrsmodelle

auf der Basis der Fahrzeuge sowie Fraktilwerte der Lasten des Eurocode 1-

Modells; Forschungsbericht Bundesanstalt für Straßenwesen

Prat M (2001) Traffic load models for bridge design: recent developments and

research. Progress in Structural Engineering and Materials 3, pp 326–334

Proske D & Hübl J (2007) Historical arch bridges under horizontal loads.

Proceedings of the 7th Arch Conference, Madeira

Proske D (2003) Ein Beitrag zur Risikobeurteilung alter Brücken unter

Schiffsanprall. Dissertation, Technische Universität Dresden

Proske D, Kaitna R, Suda J & Hübl J (2008) Uncertainty in debris flow impact

force assessment. Proceedings of the 6th International Probabilistic Workshop,

CA Graubner, H Schmid & D Proske (Eds), Darmstadt

References 97

Messungen von Spannstahlspannungen. Bauingenieur 61, pp 79–81

Quan Q (2004) Modeling current traffic load, safety evaluation and SFE analysis

of four bridges in China. Risk Analysis IV, CA Brebbia (Ed) Wessex Institute

of Technology Press, Southampton

Ril 804 (2003) Eisenbahnbrücken (und sonstige Ingenieurbauwerke) planen,

bauen und instand halten. Deutsche Bahn AG

Ril 805 (1999) Tragsicherheit bestehender Eisenbahnbrücken

Robinson JI, Prentice DJ & Ponniah D (1998) Thermal effects on a masonry arch

bridge investigated using ABAQUS. Arch Bridges: History, Analysis,

Assessment, Maintenance and Repair. Proceedings of the Second

International Arch Bridge Conference, A Sinopoli (Ed), Balkema, Rotterdam

1998, pp 431–438

Schlegel R, Will J & Popp J (2003) Neue Wege zur realistischen

Standsicherheitsbewertung gemauerter Brückenviadukte und Gewölbe unter

Berücksichtigung der Normung mit ANSYS. 21. CAD-FEM User’s Meeting

2003, Potsdam

Schütz KG (1991) Verkehrslasten für die Bemessung von Straßenbrücken,

Bauingenieur 66, pp 363–373

Soukhov D (1998) The Probability Distribution Function for Snow Load in

Germany. LACER No. 3, 1998, pp 263–274

Spaethe G (1977) Beanspruchungskollektive von Straßenbrücken, Die Straße 17,

Heft 4, pp 241–246

Thürmer E (1995) Verbreiterung einer Bogenreihenbrücke im Zuge der

Autobahn A4 bei Hainichen. 5. Dresdner Brückenbausymposium, Fakultät

Bauingenieurwesen, Technische Universität Dresden, pp 37–54

Tobias DH, Foutch DA & Choros J (1996) Loading Spectra for railway bridges

under current operation conditions. Journal of Bridge Engineering, 1 (4),

pp 127–134

UIC-report 702 (1974) Lastbild für die Berechnung der Tragwerke der

internationalen Strecken, Internationaler Eisenbahnverband UIC

Vega T (1996) Bridge management at RENFE. Recent Advances in Bridge

Engineering. Evaluation, Management and Repair, JR Casas, FW Klaiber &

AR Marí (Eds), Proceedings of the US-Europe Workshop on Bridge

Engineering, organized by the Technical University of Catalonia and the Iowa

State University, Barcelona, 15–17 July 1996, International Center for

Numerical Methods in Engineering CIMNE, pp 115–123

Vockrodt H-J (2005) Instandsetzung historischer Bogenbrücken im Spannungsfeld

von Denkmalschutz und modernen historischen Anforderungen. 15. Dresdner

Brückenbausymposium, Technische Universität Dresden, pp 221–241

Vrouwenvelder ACWM & Waarts PH (1993) Traffic loads on bridges. Structural

Engineering International, 3, pp 169–177

Weber WK (1999) Die gewölbte Eisenbahnbrücke mit einer Öffnung.

Begriffserklärungen, analytische Fassung der Umrisslinien und ein

erweitertes Hybridverfahren zur Berechnung der oberen Schranke ihrer

Grenztragfähigkeit, validiert durch einen Großversuch. Dissertation, Lehrstuhl

für Massivbau der Technischen Universität München

98 2 Loads

Brückenpfeilern, Bauingenieur, 55, pp 285–288

Wiese H, Curbach M, Al-Jamous A, Eckfeldt L & Proske D (2005) Vergleich des

ETV Beton und DIN 1045-1, Beton- und Stahlbetonbau, 100, pp 784–794

3 Computation of Historical Arch Bridges

3.1 Introduction

reaction of the arch bridge if applied to the bridge. This reaction is either

admissible or inadmissible. Inadmissible reactions of bridges include dam-

ages or failure of the bridge. To prevent such effects, the reaction of

bridges exposed to loads is usually studied in advance.

Such studies or structural analysis requires the development of an ap-

propriate numerical model.

These models usually reflect the knowledge of humans about the struc-

ture. This fact becomes especially visible for structure types with a long

history, such as arch bridges. Here, completely arbitrary models are mod-

els from Von Leibbrand (1897), Haase (1899), Gilbrin (1913), Fain

(1953), Wolf (1989), Lachmann (1990), and Falter (1998). Of course, this

chapter gives a much more detailed and structured list about certain types

of models. However, a simplified rule for choosing an appropriate model

cannot be given. Even very simple empirical rules have shown to be a sol-

id basis for bridges with ages of centuries and millenniums.

The choice of the model or model type depends on the respective ques-

tion and the provided resources. During the COST-345 (2004) report for

the EU commission, different description classes for structures and their

applications were discussed. Table 3-1 lists different levels of assessment.

This chapter starts with a discussion of simple empirical rules and ad-

vances to the latest numerical models in terms of finite element and dis-

crete element models. Many further recommendations about the different

levels of bridge assessment can be found in literature (Schueremans et al.

2003, Diamantidis 2001, Rücker et al. 2006, Enevoldsen 2008, Brühwiler

D. Proske, P. van Gelder, Safety of Historical Stone Arch Bridges, DOI 10.1007/978-3-540-77618-5_3,

© Springer-Verlag Berlin Heidelberg 2009

100 3 Computation of Historical Arch Bridges

2008, O’Connor et al. 2008, Jensen et al. 2008, Ruiz et al. 2008, and SIA

269 2008).

Table 3-1. Analysis methods recommended for each level of assessment (COST-

345 2004)

Structure Subtype Description level

type

1 2 3 4 5

Empirical or two-dimensional-model,

Two- or three-dimensional, linear-

Two- or three-dimensional, linear

the structure being assessed not

structure interaction, cracking,

or nonlinear, elastic or plastic,

elastic or nonlinear, elastic or

plastic, allowing for cracking

linear-elastic arch frame

probabilistic models

Arch bridges

properties

Bridges

In 1717, Gauter listed the following five tasks during the design process of

natural stone arch bridges (taken from Heyman 1998):

• Choice of the shape of the arch

• Choice of the arch thickness at the key

• Choice of the thickness of the foundation and abutment

• Choice of the thickness of the piers depending on the design of the arch

• Choice of the thickness of the wing walls

Over the building time of arch bridges with approximately 2000 the

design of arch bridges has been mainly carried out using empirical models.

Empirical methods are here understood as methods that describe the ul-

timate load-bearing behaviour of arch bridges based on simple geometrical

rules for the design of the arch elements. Such a simple rule is shown in

3.2 Empirical Rules 101

bearing behaviour of arch bridges are given by Corradi (1998) based al-

ready on some theoretical considerations (Table 3-2). Such empirical rules

were also introduced for the abutment, as the Blondel rule (Straub 1992)

proves. However, the foundations should not be of interest here. The major

interest here is the description of arches using empirical rules.

Fig. 3-1. Empirical rules for the design of arch bridges from Alberti around 1,450

(taken from Heinrich 1983)

Table 3-2. Types of failure for circular arches (Corradi and Filemio 2004)

Without backfill Declining backfill Horizontal backfill

Types of failure

Hinges

μ ≥ 0.395 μ ≥ 0.511 μ ≥ 0.258

K crit = 1.1136 Kcrit = 1.184 K crit = 1.044

α = 54° α = 50° α = 68°

102 3 Computation of Historical Arch Bridges

Hinges and sliding

0.309 ≤ μ < 0.395 0.406 ≤ μ < 0.511 0.236 ≤ μ < 0.258

1.2205 ≤ K crit < 1.1136 1.264 ≤ K crit < 1.184 1.1138 ≤ K crit < 1.044

29° ≤ α < 54° 20° ≤ α < 50° 44° ≤ α < 68°

Sliding

μ < 0.309 μ < 0.406 μ < 0.236

K crit = 1.2205 K crit = 1.264 K crit = 1.1138

α = 29° α = 20° α = 44°

R = Radius extrados, r = Radius intrados, µ = Coefficient of friction, α = Angle

collapse point, H = Normal force at crown, W = Self weight

R

K crit =

r

Major parameters for the description of the shape of the arch are the span l

and the rise f. They both form a ratio of f/l, which is heavily used for a first

characterisation of the arch shape. Purtak (2004) has published a statistic

about this ratio (Fig. 3-2).

Fig. 3-2. Ratio of arch rise f to arch span l related to the span l (Purtak 2004)

3.2 Empirical Rules 103

f f (3-1)

s= =

l 2⋅a

Full circular arches have an s of 1/2. For segmental arches, s reached

values between 1/6 and 1/9. Sometimes additionally a minimum angle of

60° at the springing was required that corresponded to an s of 1/7.5. At the

end of the 18th century, the basket arch became more and more popular.

At the Neuilly Bridge, Perronet used 11 circle segments to shape the arch.

However, often lower numbers of circles were used such as three, five, or

seven. Perronet’s Bridge reached s = 1/4. Also, for the latter, ellipses

s-values of 1/4 were used, or in general greater than 1/5 were used. How-

ever, in contrast to such rules, bridges with much lower s values were also

constructed, such as the Nemours Bridge by Perronet in 1792 with s = 1/10

or s = 1/15 (Souppes) (Corradi and Filemio 2004, Corradi 1998).

Additionally to the choice of the s value, the number of circle elements

in an arch was ruled for basket arches. For example, if the s was equal to

1/3 and the span of the arch was higher than 10 m, then three circular ele-

ments were suggested. Between 10 and 40 m span, five circular elements

were recommended, and for a span greater than 40 m, about seven seg-

ments should be chosen. For s = 1/4, the number of circular segments

should then be five, seven, and nine. For even greater s values, a circle

with a radius r of

l2 + 4 ⋅ f (3-2)

r=

8⋅ f

was suggested (Corradi and Filemio 2004).

However, the low number of circular segments in basket arches yielded

to an aesthetical problem of the transition of the different segments.

Therefore, later recommendations gave a higher number of segments.

Also, parallel to the basket arches as mentioned before, alternative shape

functions were chosen: Sejourne recommended ellipses and Viviani used a

cycloid for the arch shape of a bridge over the river Arzana. At two bridges

between Pistoia and Modena, a circular evolvente was used for the arch

shape. Additionally, during the medieval ages, often ogival designs with

parabolic functions were used. And in the 18th century, additional

catenaries were applied. For example, bridges in Boucicault, Orelans, and

Avignon used this function (Corradi and Filemio 2004, Corradi 1998).

Tolkmitt recommended the shape function subject to the loading. In

general, Tolkmitt chose the following function

104 3 Computation of Historical Arch Bridges

n (3-3)

p = ∑ ak ⋅ z 2 k

k =0

with ak as constant and z as longitudinal axis. The shape should then fol-

low the function

d2y p (3-4)

2

=− .

dz H

The solution of this differential equation is as follows:

1 n ak (3-5)

y= ∑

H k =0 (2 ⋅ k + 1) ⋅ (2 ⋅ k + 2)

⋅ k 2⋅k + 2 .

Freyssinet has used n=4 for the Bernard Bridge over the Balbigny–Regny

railway line. Tables for the computation of the shape were among others

provided by Kögler (Corradi and Filemio 2004).

A further arch shape function is given with

1 (3-6)

y= log(cos( a ⋅ z )) .

a

Lebert’s parabolic functions were

y = k ⋅ log(cos( a ⋅ z )) (3-7)

and

y = m ⋅ log(cosh( a 2 ⋅ z )) , (3-8)

Further information about the arch shapes can be found in Chapter 1.

the typical three-hinge model nowadays. Based on this model, the thickness

of the arch e subject to the extrados radius R could be estimated (Kurrer

2002)

e1 = 2 ⋅ R ⋅ (3 − 2 ⋅ 2 ) = 0.343 ⋅ R (3-9)

3.2 Empirical Rules 105

r = R ⋅ (4 ⋅ 2 − 5) = 0.657 ⋅ R . (3-10)

The line of thrust is therefore completely inside the arch. However, the

arch thickness can further be computed for a failure angle of the arch. For

example, if 45° are assumed, then the thickness becomes

eo. Fabri = 0,5 ⋅ R ⋅ (2 − 2) = 0.293 ⋅ R . (3-11)

lower thickness can also be stable, but that depends on the special condi-

tions (Kurrer 2002).

In 1730 Couplet introduced a thickness value based on the following

observations:

eu.Couplet = 0,096 ⋅ R . (3-12)

As mentioned before, the thickness depends on the assumed failure angle

in the arch. Whereas Fabri and Couplet assumed 45°, Heyman estimated

the angle with 58.8° and computed the required thickness (Kurrer 2002) as

follows:

eu , Heyman = 0,101 ⋅ R . (3-13)

crown (Kurrer 2002):

2 ⋅r (3-14)

eu , Alberti = = 0,131 ⋅ R .

10

Croizette-Desnoyer’s formulae is another empirical rule for thickness

(Corradi and Filemio 2004, Martín-Caro and Martínez 2004, Modena et al.

2004)

e = a + b ⋅ 2⋅r (3-15)

road traffic). Genio Civile suggested the following formulae (Corradi and

Filemio 2004):

l (3-16)

e = 0.05 ⋅ h + 0.40 ⋅ l + 2 ⋅ (10 + l ) ⋅ .

100 ⋅ f

Sejourne has also suggested a formulae (Martín-Caro and Martínez

2004). Busch and Zumpe (1995) have investigated historical arch bridges

and have given the following rule for the arch thickness at the crown:

106 3 Computation of Historical Arch Bridges

This value can additionally be adjusted between 0.3 and 1.9 depending

on the masonry quality, the coverage and the loading type. The minimum

coverage thickness can also be computed by

h = 0.5 + 0.015 ⋅ l . (3-18)

The axial force N in the arch at the crown is given by

q ⋅ l2 (3-19)

N=H=

8⋅ f

where q represents the loading.

The axial force can be used to compute the stress inside the arch, which

can then be compared to the compress strength of the arch. A simplified

rule for stress evaluation at an arch crown was given by Dischinger (1949):

l2 (3-20)

σ = γi ⋅ ,

8⋅ f

where γi is the characteristic weight of the bridge. For long span, low-

pitched arches, γi amounts to 30 kN/m and for slender, tall bridges, the

3

3

value reached 40 kN/m .

How the rules for the thickness of the arch key or arch crown were used

in practice is shown by some statistical investigations from Purtak (2004)

and Busch and Zumpe (1995). Figures 3-3 and 3-4 show the results of

these investigations in terms of key thickness versus span.

3.2 Empirical Rules 107

Fig. 3-3. Ratio of arch thickness at crown to span according to Purtak (2004)

Fig. 3-4. Ratio of arch thickness at crown to span according to Busch and Zumpe

(1995)

Besides the thickness of the arch at the crown, in many cases, simply the

key stone thickness e versus the span l was given. Alberti gave e/l = 1/15,

108 3 Computation of Historical Arch Bridges

Palladio 1/12, and Serlio 1/17. These values were mainly based on obser-

vations of Roman bridges. For example, the Pont de Gard reached e/l = 1/15.

Gautier recommended

1⋅ l2

e= (3-21)

18

at the beginning of the 18th century for bridges made of strong stones

and with a span greater than 10 m. For bridges with softer stones, a mini-

mum thickness of 0.32 m was required.

Many further rules were established, most of them in the 19th century,

and are listed in Table 3-3 and visualised in Fig. 3-5 (Corradi and Filemio

2004, Corradi 1998).

Dupuit for segmental arches e = 0.15 ⋅ l 0.5

Dupuit for semi-circular arches e = 0.20 ⋅ l 0.5

Rankine e = 0.191 ⋅ R 0.5

Gautier for semi-circular arches e = 0.32 + 1/15 ⋅ l

Perronet for semi-circular arches e = 0.325 + (1/ 24 − 1/144) ⋅ l

Lesguillier for semi-circular arches e = 0.10 + 0.20

Dejardin for semi-circular arches e = 0.30 + 0.045 ⋅ l

Dejardin for circular arches e = 0.30 + 0.025 ⋅ l

Dejardin for elliptical arches e = 0.30 + 0.014 ⋅ l

Further equation e = 0.2 ⋅ l

German and Russian engineers for segmental e = 0.43 + 0.1 ⋅ ρ

arches

Perronet e = 0.325 + 0.0694 ⋅ ρ

Perronet for semi-circular arches e = 0.325 + 0.035 ⋅ l

Lesguillier for segmental arches e = 0.10 + 0.20 ⋅ l 0.5

L’Eveillé for segmental arches e = 0.33 + 0.033 ⋅ l

German and Russian engineers for semi-circular e = 0.43 + 0.05 ⋅ l

arches

Gauthey for semi-circular arches e = 0.33, l < 2 m

e = 0.33 + 0.020833 ⋅ l,

Gauthey for semi-circular arches

2 m < l < 16 m

e = 0.0416 ⋅ l,

Gauthey for semi-circular arches

16 m < l < 32 m

e = 1.33 + 0.020833 ⋅ (l − 32m),

Gauthey for semi-circular arches

l > 32 m

E. Roy for semi-circular arches e = 0.30 + 0.04 ⋅ l

Michon for semi-circular arches e = 0.40 + 0.04 ⋅ l

R – radius of the circle passing through the crown joint and the intrados springing in

metre, l – span in metre, e – key stone thickness in metre and ρ as curvature radius,

3.2 Empirical Rules 109

l

for segmental arch bridges, ρ = ,

2

l2 f

for full circular arch bridges, ρ = +

8⋅ f 2

l2

and for elliptical arch bridges, ρ =

4⋅ f

Fig. 3-5. Crown joint thickness e versus the span l according to different functions

(Huerta 2004, Corrodi 1998)

Castigliano detected that the equations from circular arches and segmen-

tal arches can be converted by considering a limited span of the circular

arch in terms of l’=0.866·l. Using this equation, the factor for Dupuits for-

mula changes from 0.2 to 0.14. The conversion was also valid for

L’Eveillé’s equation (Coraddi 1998).

Kaven’s equation was given as

⎛ l⎞ (3-22)

e = 0.25 m + l ⋅ ⎜ 0.025 + 0.00333 ⋅ ⎟ .

⎝ f⎠

The equation was valid for spans smaller than 12 m and arches made of

very strong stone material. Furthermore, the height of the superstructure

110 3 Computation of Historical Arch Bridges

above the crown was limited to 1.5 m. Greater fillings required a specific

adaptation

Road bridges: 1 + 0.214 ⋅ h , (3-23)

Additionally, if the bridge was made up of brick stone, the arch thickness

should be further increased by a factor of 1.5.

Huste’s equation was given as

e =α⋅ ρ (3-24)

with

α = 0.165 for arches built of strong stone material,

α = 0.220 for arches built of brick stones,

α = 0.247 for arches built of soft stone material,

According to Corradi (1998), Italian engineers used

l 20 + l l (3-25)

e = 0.20 m + + + .

40 1000 f

Résal’s equation is given as (Coraddi 1998)

l (3-26)

e = 0.15 m + 0.20 ⋅

2⋅ f

Croizette-Desnoyers’s equation is given as (Corradi 1998)

e = k1 + k2 ⋅ ρ (3-27)

with ρ as curvature of the arch. The factors depend on the shape and the

type of the bridge, and are given in Table 3-4 (Corradi 1998).

Circular arch

Road bridge Railway bridge

f/l k1 k2 k1 k2

1/2 0.15 0.15 0.20 0.17

Segmental arch

Road bridge Railway bridge

f/l k1 k2 k1 k2

1/4 0.15 0.15 0.20 0.17

1/6 0.15 0.14 0.20 0.16

3.2 Empirical Rules 111

Segmental arch

Road bridge Railway bridge

1/8 0.15 0.13 0.20 0.15

1/10 0.15 0.12 0.20 0.14

1/12 0.15 0.12 0.20 0.13

For skewed bridges, Résal recommended increasing the thickness ac-

cording to

e (3-28)

e' = ,

sin α

where α is the skewness of the bridge in grad.

Further equations are given in Leliavsky (1982). Taken from that refer-

ence, the equation of Heinzerling is given as

e = k1 + k2 ⋅ r . (3-29)

The factors are given in Table 3-5.

k1 k2

0.4 0.028 Brick stone masonry

0.42 0.032 Ashlar masonry

1 l ⋅Q (3-30)

e = 0.2 + ⋅ .

21 zul σ

Further equations from Melan, Landberg and Mehrtens should only be

mentioned (Leliavsky 1982).

In 1845, Dejardin suggested the following formula for the thickness of

arches:

e (3-31)

e1 =

cos ϕ

The angle is measured from the crown (Corradi 1998). If this formula is

applied, then the extrados follows a Nicomede conchoids or Pericycloid if

the intrados follows a circular shape. However, the suggested formula can

not be applied for half-circle arches because the thickness of the stones at

springing becomes infinite. Therefore, the thickness of the arch stones is

often not altered for an angle of 60°.

112 3 Computation of Historical Arch Bridges

e1 = 1.4 ⋅ e . (3-32)

However, in contrast to the description of the change of the stones, he fur-

thermore suggests (Corradi 1998)

e (3-33)

e1 = .

cos ϕ

Another equation is given as

1.125 ⋅ ( ρ + e ) ≤ R < 1.25 ⋅ ( ρ + e ) . (3-34)

The results of the later equation give values between Dejardin’s and

Tavernier’s suggestions. Furthermore, Croizette-Desnoyers’s rules were

widely applied:

f f (3-35)

s= = .

l 2⋅a

The parameters for the formula are given in Table 3-6.

s 1/3 1/4 1/5 1/6 1/8 1/10 1/12

Half circular - 1.80 - 1.40 1.25 1.15 1.10

Elliptic 1.80 1.60 1.40 - - - -

bilinear.

Besides the design of the arch shape, the thickness of the arch itself at the

crown and at the springing of the abutment has to be chosen carefully.

Many empirical relationships were published concerning this question.

Design variables were Ss as the thickness of the abutment, r as the rise of

the arch, L as width, e as the thickness of the crown joint, h as the distance

from the springing line to the foundation base, and h1 as the height of the

backing. The parameter H is defined as

H = h + f + e + 0.60 , h1 < 0.60 m , (3-36)

H = h + f + e + h1 , h1 > 0.60 m .

3.2 Empirical Rules 113

abutment (Table 3-7).

Table 3-7. List of equations for the computation of the abutment thickness Ss

(Corradi 1998)

Author Equation

Semicircular arch

Lesguillier Ss = (0.60 + 0.04 ⋅ h ) ⋅ l

L’Éveillé 0.865 ⋅ l ⋅ ( h + 0.25 ⋅ l )

Ss = (0.60 + 0.162 ⋅ l ) ⋅

H ⋅ (0.25 ⋅ l + e )

German and Russian 5 h h

engineers Ss = 0.305 + ⋅l + + 1

24 6 12

Segmental arch bridges

Lesguillier ⎡ ⎛l ⎞ ⎤

Ss = ⎢0.60 + 1.10 ⋅ ⎜ − 2 ⎟ + 0.04 ⋅ h ⎥ ⋅ l

⎣ ⎝ f ⎠ ⎦

L’Éveillé l⋅h

Ss = (0.33 + 0.212 ⋅ l ) ⋅

H ⋅ ( f + e)

German engineers ⎛ 3 ⋅ l − f ⎞ 2 ⋅ h + h1

Ss = 0.305 + 0.125 ⋅ l ⋅ ⎜ ⎟+

⎝ l+ f ⎠ 12

Italian engineers ⎛ 10 + 0.5 ⋅ l ⎞ ⎛ l ⎞

Ss = 0.05 ⋅ h + 0.20 ⋅ l + ⎜ ⎟⋅⎜ ⎟

⎝ 100 ⎠ ⎝ f ⎠

Semi-elliptical arches

Lesguillier ⎡ ⎛l ⎞ ⎤

Ss = ⎢0.60 + 0.05 ⋅ ⎜ − 2 ⎟ + 0.04 ⋅ h ⎥ ⋅ l

⎣ ⎝f ⎠ ⎦

L’Éveillé

⎛ h + 0.54 ⋅ f ⎞ ⎛ 0.84 ⋅ l ⎞

Ss = (0.43 + 0.154 ⋅ l ) ⋅ ⎜ ⎟⋅⎜ ⎟

⎝ H ⎠ ⎝ 0.65 ⋅ f + e ⎠

German engineers ⎛ 10 + 0.5 ⋅ l ⎞ ⎛ l ⎞

Ss = 0.05 ⋅ h + 0.20 ⋅ l + ⎜ ⎟⋅⎜ ⎟

⎝ 100 ⎠ ⎝ f ⎠

Manuale del’Ingenere ⎛ 0.17 ⋅ l ⎞

Ss = l ⋅ ⎜ 0.42 + + 0.44 ⋅ h ⎟ bei h1 < 1.50 m

⎝ 2 ⋅ f ⎠

⎛ 0.17 ⋅ l ⎞

Ss = l ⋅ ⎜ 0.42 + + 0.44 ⋅ h ⎟ + 0.0185 ⋅ H ⋅ h1

⎝ 2 ⋅ f ⎠

bei h1 ≥ 1.50 m und mit H = h + f + e

114 3 Computation of Historical Arch Bridges

Author Equation

Hütte l ⎛ 3⋅ l − f ⎞ h

Ss = ⋅ ⎜ ⎟ + 1.00 +

8 ⎝ f +l ⎠ 6

and for semi-circular arches

5 h

Ss = ⋅ l + 1.00 +

24 6

Croizette-Desnoyers ⎛ l⋅h ⎞

Ss = 0.33 + 0.212 ⋅ l ⋅ ⎜ ⎟

⎝ H ⋅ ( f + e ) ⎠

Italian railway engineers Ss = 0.20 + 0.030 ⋅ ( ρ + 2 ⋅ e) + 0.10 ⋅ h

Besides the abutment thickness, the pier width also had to be estimated.

Here, some general rules such as the pier width have to have at least the

thickness of the abutment, and the pier width has to have at least twice the

thickness of the crown joint to be fulfilled. Furthermore, possible hydro

dynamic water pressure had to be considered. Finally, some equations for

the computation of the pier width are given in Table 3-8. The variables are

h as the height from the foundation springing line to the haunch joints of

rupture, H represents the distance between the foundation line and the

bridge road line, e is the thickness of the crown joint, f is the distance be-

tween the crown joint extrados and the haunch joint of rupture, d is the

horizontal distance between the haunch joints of rupture, P is the weight of

the section of arch up to the joint of rupture, ω is the weight of the bridge

construction material, and k is a safety factor (Corradi 1998).

Table 3-8. Computation of the pier width according to different authors (Corradi

1998)

Author Formula and arch shape

L’Éveillé Circular arch

h

E = (0.33 + 0.212 ⋅ D ) ⋅ H

f +e

D

Half-circular arch

h + 0.25 ⋅ D

E = (0.60 + 0.162 ⋅ D) ⋅ H

0.25 ⋅ D + e

0.865 ⋅ D

Basket arch

3.2 Empirical Rules 115

h + 0.54 ⋅ D

E = (0.42 + 0.154 ⋅ D ) ⋅ H

0.465 ⋅ D + e

0.84 ⋅ D

Séjourne 1

E > ⋅l

5

E = 0.4 + 0.15 ⋅ l

E = 0.8 + 0.1 ⋅ l

Perronet E = 2.25 ⋅ e

Castigliano S

E= s

2

Colombo E = 0.20 ⋅ h '+ 0.6

1 1

E = ⋅l , E = ⋅l

6 10

h’ as pier height from foundation to springing

Further applied formula E = 0.292 + 2 ⋅ e

Further applied formula E = 2.50 ⋅ e für l ≤ 10 m

E = 3.50 ⋅ e für l > 10 m

Further applied formula 1 1

⋅l < E < ⋅l

10 5

Minimum value E > ∑ e1 , if several arches

Weber (1999) has estimated the maximum arch length of certain materials

based on the breaking length of the material. The breaking length of the

material is the maximum length of a bar of a certain material where the

tensile or compression strength of the material and the stress caused by the

self-weight of the bar are equal. In Table 3-9, the breaking length for

different construction materials is summarized.

σ (3-37)

lim l = π ⋅ .

γw

Table 3-9. Theoretical maximum span of arch bridges for different materials

(Weber 1999)

Material lim l in m

St 37 6,467

GS-52 7,585

116 3 Computation of Historical Arch Bridges

Material lim l in m

Brick masonry 205–748

Natural stone masonry 145–785

Reinforced concrete 228–1870

Lightweight concrete 524–2,244

Coniferous wood 6,912–8,639

Very interesting parameters describing the capacity utilization of an arch

bridge are so-called coefficients of boldness. Such boldness numbers were

developed by different authors. First, the number originating from Résal is

introduced. The number is computed as

l⋅ρ (3-38)

A= .

2

The variables are explained in Table 3-10 and in the footnotes of Table 3-3.

Furthermore, examples for different bridges are given there.

Bridge Span l in m Rise f in m Arch thickness e in cm A

Fouchard Bridge 26.0 2.6 33.8 439

Nogent Bridge 50.0 25.0 25.0 625

Tournon Bridge 49.2 17.7 25.9 639

Antoinette Bridge 50.0 15.9 27.6 690

Lavaur Bridge 61.5 27.5 30.9 931

Moskwa Bridge 44.8 5.6 7.6 1066

Cloris Bridge 50.0 7.4 46.0 1150

Chester Bridge 61.0 12.8 42.9 1308

Cabin John Bridge 67.0 17.6 40.7 1363

Trezzo Bridge 72.3 20.7 42.0 1517

Soupees arch 37.9 2.1 88.0 1667

boldness:

l2 (3-39)

k= .

f

Spangenberg’s number is widely known. However, Weber (1999) has

shown that for a parabolic arch second order, Spangenberg’s number cor-

responds with eight times the curvature radius of the arch at the crown.

3.2 Empirical Rules 117

The number can also be adapted to other arch curvatures as shown in Table

3-11.

Arch shape Spangenberg’s number represents

Parabola second order 8⋅ ρ

Parabola fourth order 8 ⋅ ρ ⋅ (1 − c )

⎛ l2 ⎞

Catenary First type ≈ 8 ⋅ ρ ⋅ ⎜1 − 2 ⎟

⎝ 48 ⋅ ρ ⎠

⎛ f ⎞

Segmental circular arch 8 ⋅ r ⋅ ⎜1 − ⎟

⎝ 2 ⋅r ⎠

Sinus curvature (half wave) π2 ⋅ρ

Upper elliptic half 4⋅ρ

l arch span, f arch rise, ρ curvature radius

number. For example, the consideration of the arch thickness would be

compelling. Based on the Hugi’s self-capacity utilization

g (3-40)

ω= .

g+ p

Weber (1999) defines

g χ (3-41)

χ= and ω = .

p 1+ χ

Weber (1999) has then introduced a boldness number that considers the

• Static system

• Cross-section geometry

• Load pattern

• Construction material

• Slenderness

• Structural size

The number is given as

3 1 (3-42)

⋅λ +

2 k0

ω=

3 1

16 ⋅ θ + ⋅ λ −

2 k0

118 3 Computation of Historical Arch Bridges

with

f l l f (3-43)

k0 = ,θ= ,λ= , l = d

l l d4 γw

Table 3-12.

For further analytical models of the ultimate load-bearing behaviour, see

Audenaert et al. 2007.

Table 3-12. Weber’s ω and Spangenberg’s boldness number k for certain bridge

structures including further parameters (Weber 1999)

Bridge1 Year l in m f/l d4 l/d4 M.2 ω k

Thur Bridge Kurmmenau (S) 1911 63.3 1/4.57 2.26 27.99 LM 0.65 289

Iller Bridge St. Buchlow-Kempten 1906 63.8 1/2.48 1.94 32.89 TC 0.51 158

Iller Bridge St. Buchlow-Lindau 1906 64.5 1/2.34 1.98 32.58 AM 0.51 151

Prut Bridge Karemča (U) 18923 65.0 1/3.63 2.60 25.0 AM 0.48 236

Rummel Bridge Sidi-Rached (A) 1912 68.0 1/3.09 1.52 44.74 AM 0.65 210

Adda Bridge bei Morbegno (I) 1904 70.0 1/7.00 1.73 40.46 AM 0.61 490

Roizonne Bridge, La Mure (F) 1916 79.5 1/2.10 1.60 49.66 AM 0.70 169

Valserine Bridge, Montages (F) 1910 80.3 1/4.01 2.00 40.15 AM 0.69 322

Pétruss Bridge (L) 1903 84.6 1/2.71 1.80 47.03 AM 0.71 229

Soča Bridge, Solkan (SL) 1927 85.0 1/3.90 2.45 34.69 AM 0.59 332

Friedens Bridge in Plauen 1904 90.0 1/5.00 1.65 54.55 LM 0.69 495

Lot Bridge in Villeneuve (F) 1920 96.3 1/6.23 1.45 66.38 TC 0.82 600

Nanpan Bridge Changhong (C) 1961 112.5 1/5.43 2.21 50.90 AM 611

Chang Bridge Jiuxigou, Feng.(C) 1972 116.0 1/80 1.87 62.03 AM 928

Wuchao Bridge, Fong-Huan (C) 1990 120.0 1/50 1.50 80.00 AM 0.88 600

120 m Bridge Project (A-H) 120.0 1/3.6 5.10 23.53 AM 0.69 432

1

S – Switzerland, SL – Slovenia, U – Ukraine, C – China, A–H – Austria–Hungry,

L – Luxemburg, F – France, A – Algeria, names without country identification are

in Germany located

2

LM – Layered masonry, TC – tamped concrete, AM – Ashlar masonry,

M. – Material

3

1916 destroyed by the Russian army, 1927 reconstructed

3.2 Empirical Rules 119

probably the best known simple method for the load-bearing assessment of

historical arch bridges. The method is heavily applied not only in Great

Britain (Huges and Blackler 1997), the country of origin, but also in many

other countries since it was included in UIC-Codex (1995).

Early works for the development of this method were carried out by

Pippard in the 1930s. The basic assumption was linear-elastic behaviour of

the material. Especially during the World War II, Pippard’s method found

wide application for the assessment of historical arch bridges under mili-

tary loads. However, since military loads changed later due to the new

NATO, the method had to be revised. The revision based on the work by

Pippard yielded to the first-generation MEXE method. The testing of sev-

eral arch bridges in Great Britain in the 1950s provided newer results,

which could be considered again for a revision of the method. Therefore,

in the 1960s, also caused by changing live loads, the method was modified

(Das 1995).

The application of the method is simple and fast. An example will show

that. First, the conditions found have to comply with the boundary condi-

tions of the method. The boundary conditions are given in Table 3-13.

Requirements Example

Span smaller then 20 m 10 m

Rise greater then ¼ of the clear span ¼ ×10 = 2.5>1.8 m

Filling above crown is between 30 and 105 1.2–0.5 = 0.7 m

cm

Qadm = Q p ⋅ f (3-44)

with

Qadm (3-45)

qadm =

1.5 m

and

120 3 Computation of Historical Arch Bridges

1 (3-46)

f = fS ⋅ f M ⋅ f J ⋅ fC ⋅ f N ⋅ .

fΦ

The factor f considers a variety of parameters. The first one is the arch

shape factor:

rq 1.4 (rc − rq ) 0.6

(1.8 − 1.4)0.6 (3-47)

= = 0.78 → fS = 2.3 ⋅ = 2.3 ⋅ = 0.95 .

rc 1.8 rc 1.4

Further factors, which have to be taken from diagrams and tables, are

the following:

Material factor

f M = 1.0 . (3-48)

Joint factor

f J = fW ⋅ fmo . (3-49)

Joint thickness factor

fW = 0.8 → Joint thickness > 12.5 mm . (3-50)

Mortar factor

fmo = 0.9 → weak, crumbly mortar . (3-51)

Factor for the arch condition

fc = 0.85 → Longitudinal cracks in the middle third of the arch (3-52)

Factor for the numbers of arches

f N = 0.8 → Arch supported by two piers . (3-53)

And, f can be computed as

1 (3-54)

f = 0.95 ⋅1.0 ⋅ 0.8 ⋅ 0.9 ⋅ 0.85 ⋅ 0.8 ⋅ = 0.372 .

1.25

Using this value, one can read from a diagram (UIC-Codex 1995)

Q p = 425 kN (3-55)

3.2 Empirical Rules 121

0.158 MN (3-57)

qadm = = 0.105 MN/m .

1.5 m

As shown, the method can be quickly applied. This explains the wide

application. However, some authors such as Brencich et al. (2001) have

criticized the method that produces unsafe results under some circums-

tances.

Therefore, it is not surprizing that beside the successful and wide appli-

cation of the MEXE method in the last few years, major efforts have been

undertaken to develop successors of the MEXE-method. A few such me-

thods will be introduced in the following sections.

based on nearly 800 bridge models with different parameters for which the

ultimate load bearing was investigated with the FE-program Sofistic. Table

3-14 gives a summary of the investigated parameter combinations for sin-

gle loads and Table 3-15 for uniform loads. The meaning of some of the

parameters is shown in Fig. 3-6. The computation database was then used

for derivation of approximation equations. However, of course, such equa-

tions include some simplifications. For example, the live load was consi-

dered either only as single load with changing position or as a uniform

load over half entire arch span, respectively. The position of the single

load was modified between 1/5 and 1/2 of the arch span. Furthermore, for

the single load, only the development of a four-hinge mechanism or shear

failure was considered as failure. For the uniform load, additional com-

pression failure was allowed.

Spannweite l in m c/l f/l ho in m hp in m bp/l µ γ in γfill in

kN/m3 kN/m3

5.00 0.10 1/2 0.25 2.0 1/4 0.60 20.0 18.0

7.50 0.09 1/4 0.50 5.0 1/6 0.80

10.00 0.07 1/6 2.00 10.0 1/8

12.50 0.06

15.00 0.05

17.50

20.00

122 3 Computation of Historical Arch Bridges

Spannweite l in m f/l ho in m fc in MPa γ in kN/m3

5.00 1/2 0.40 4.0 20.0

10.00 1/6 6.0

20.00 8.0

10.0

The arch width was chosen constant with 3.0 m, and the load spreading

in the backfill was considered with 30°. Dead load was included with a

partial safety factor of 1.0, and uniform live load was considered with a

partial safety factor of either 1.0 or 1.35. Further assumptions are the

following:

• The bridge is straight and neither skewed nor bended

• The bridge is only a one span arch bridge

• Adjacent arches have the same span

• Abutments are considered as infinite stiff

• The bridge is not damaged

• The backfill is not hollowed

• The span is between 2 and 20 m

• The minimum ratio of rise to span is f/l>1/6

• The ratio of arch thickness at the crown should follow the values given

in Table 3-15

• Backfill exists above extrados between the abutments. The height of the

backfill is computed based on the ratio of rise to span for high arches

(f/l = 1/2) with hbackfill = 0.6 × f and for flat arches (f/l = 1/6) with

h = 0.3 × f

• The filling above crown is assumed between 0.25 and 2.0 m

• The maximum pier height is considered with 10 m

• The maximum pier width is at least 1/6 of the arch span

3.2 Empirical Rules 123

Table 3-16. Ratio between arch thickness at crown c and arch span l

l in m 2.0–5.0 5.0–7.5 7.5–10.0 10.0–15.0 15.0–20.0

c/l ≥ 0.10 0.09 0.07 0.06 0.05

Currently, Harvey (2007) and Harvey et al. (2007a) have developed a

substitute method for the MEXE method. Unfortunately, not much was

known about the new method at the time this book was printed. However,

new material will probably be published in the years to come.

In general, Harvey et al. (2007) promised that the method will be simp-

ler to use and will behave robustly. The method can be applied by artisans

and will probably include nomograms such as shown in Fig. 3-7.

Fig. 3-7. General outline of the Harvey method (Harvey et al. 2007)

Purtak et al. (2007) have developed curvatures for the ultimate load of his-

torical arch bridges based on finite element simulations using the program

ANSYS. The simulation considers not only material nonlinear behaviour

of the stone material using the Mohr–Coulomb model, but also openings of

joints. The load-bearing curvatures have been constructed for a great vari-

ety of different materials and geometrical parameters, such as stone width

to height, joint thickness, stone and mortar compression strength, stone

tensile strength, and eccentricity. Due to the variety of different parameters,

the curvatures are able to cover a wide range of different conditions, such

as compression failure or development of mechanisms. Figure 3-8 shows

only a simple example of such load-bearing curvatures.

124 3 Computation of Historical Arch Bridges

Fig. 3-8. Example of ultimate load versus eccentricity functions (Purtak et al.

2007)

Martinez et al. (2001) have also developed ultimate load-bearing

curvatures for the crown arch thickness depending on the stone material’s

compression strength, the ratio between rise and span, and the span. Again,

this method permits a very quick estimation of the ultimate load-bearing

behaviour of the arch bridge.

Aita et al. (2007) discuss the application of two different methods for the

assessment of stress levels in arch systems based on geometrical and mate-

rial properties. They apply the methods to semicircular and pointed arches

and receive the so-called stability area curvatures. Such stability area cur-

vatures can be extended towards diagrams including maximum arch-wall

system height related to arch thickness or maximum height versus volume.

estimating the load-bearing capacity for arch bridges in a fast way. How-

ever, it has been strongly suggested that a computer program should be

provided for the computation since several correction parameters consider-

ing different conditions at the bridge have to be estimated.

3.3 Beam Models 125

computed tables comparable to the technique by Purtak. Finally, the load-

bearing capacity of the arch bridge can be related to diagrams shown in

Fig. 3-9 (Heinlein 2008, Pauser 2005).

Fig. 3-9. Interaction diagram for an arch cross section (Heinlein 2008)

With the introduction of analytical models, the epoch of beam models was

also started. Whereas the first beam models had to comply with the re-

requirement of simple hand computation (static determined structures), this

requirement does not hold nowadays. With the loss of this requirement,

beam models were more and more extended to consider more and more

effects on the arch bridge load behaviour yielding more precise results.

This evolution is shown in Table 3-17. Whereas the first models simply

chanced the number of predefined hinges, the backfill, elastic foundation,

and roadway structures were also considered later. The last model from

Gocht (1978) is mainly suited for arch bridges with a reinforced roadway

or railway slab.

126 3 Computation of Historical Arch Bridges

Type of model Visualization

Fixed arch without

backfill

backfill

backfill according to Ril

805 or UIC Codex

Backfill considered

Model according to

Voigtländer (1971)

Model according to

Model (1977)

roadway slab according

to Gocht (1978)

the arch itself, although the biggest single contribution to the load-bearing

behaviour of arch bridges, is only one element of several. Weber (1999)

has quantified the contribution of the single elements of an arch super-

structure based on a test. In general, he proposed a factor of 1.5 increased

load-bearing capacity from the single-span pure arch to the single-span

3.3 Beam Models 127

arch bridge. Based on Fig. 3-10, this factor is wide on the safe side. Another

rough measure considering further elements in an arch superstructure was

given by the former German national railway. Without any further proof

the railway permitted an increase of 20% in the load-bearing capacity of

historical arch bridges by building reinforced concrete slabs on the bridges

(DR 1985). This corresponds with the model from Gocht (1978).

The load-increasing effects of roadway and pavement concrete are also

known from other bridge types. For example, Gutermann (2002) has inves-

tigated the influence of the three elements and found experimentally an in-

crease of the load-bearing capacity by the road asphalt layer of about 3%,

by the pavement concrete of about 12%, and the protection concrete up to

10%. However, the investigation of Gutermann (2002) was carried out on

serviceability levels and in terms of stresses and deformations.

Fig. 3-10. Contribution of different structural elements to the load bearing based

on measurements on the railway arch bridge Schauenstein (Weber 1999)

known for a long time. For example, Craemer (1943) already stated a

change of the line of trust caused by the backfill in arch bridges. Fischer

(1940–1942) has investigated the shear stress between the arch and the

backfill. For simplification reasons, he only considered three-hinge arches.

Jäger (1938) proved an increased load-bearing capacity of arch bridges

when shear forces were transferred to the backfill. Herzog (1962) also

showed an higher load-bearing capacity. He found a lower eccentricity

128 3 Computation of Historical Arch Bridges

when backfill was acting. However, he only considered the same Young

modulus for the arch and the backfill. Bienert (1959–1960) and Bienert et

al. (1960–1962) also discussed the effects on the load-bearing behaviour of

arch bridges caused by backfill. A very intensive list of references can be

found in Gocht (1978).

Furthermore, Gocht (1978) has developed some models that should be

shown here. First, he developed a model that if no shear forces are trans-

ferred between the arch and the backfill, a roadway slab still exists. Later,

he discussed the effects of an active height of the backfill. Figures 3-8 and

3-9 show the height considering a solid joint or a sliding joint between the

arch and the backfill. Also, Fig. 3-11 considers whether a joint at the

springing exists. Figure 3-12 shows that a solid joint between the arch and

the backfill can also have negative effects. Then, a flat arch can develop

inside the backfill. These effects also change the span of the arch.

Fig. 3-11. Model of the effective height of the arch by consideration of the backfill

according to Gocht (1978). Left figure shows the height without a springing joint

and right figure with springing joint

Fig. 3-12. Effective height for a half-cycle arch based on Gocht (1978)

Indeed, the collapse of the Italian Traversa railway bridge between Tu-

rin and Genoa was related to such a change of system (Brencich and Colla

2002). Figure 3-13 shows the plan and front view of the collapsed bridge.

Please compare this figure with Fig. 3-12 (right).

3.3 Beam Models 129

Fig. 3-13. Traversa Bridge in Italy after partial collapse according to Brencich and

Colla (2002)

Smith et al. (2004) have given an estimation of the increase of the load-

bearing behaviour by backfill and coverage, respectively, as shown in Fig.

3-14.

et al. 2004)

Molins and Roca (1998) not only have shown the influence of the

backfill, but also of other structural elements. For two arches, they have

computed the failure load to 15 and 16 kN, respectively, if only the arch

itself is considered. If the model is extended with spandrel walls, the

failure load increases to 46 and 48 kN, respectively. However, only if

tensile bars (tie) are considered, the experimental failure loads in the

range of 90 kN can be achieved by numerical investigation. That means

that the arch itself contributes only by 25%, which fits very well with the

130 3 Computation of Historical Arch Bridges

results from Weber (1999). The results from Molins and Roca (1998) are

summarized in Table 3-18.

Arch c Arch d

Ultimate load measured 60 kN 95 kN

Ultimate load computed with

arch and backfill at arch 2 15 kN 16.1 kN

Arch and spandrel walls 46 kN 48 kN

Arch, spandrel wall, and tie bars considered 60 kN 91 kN

They have investigated the influence of the backfill at the real size tests in

the United Kingdom at the Prestwood Bridge. The bridge had a span of

6.55 m, a rise of 1.428 m, and the thickness of the arch at the key of 0.22

m. The coverage at the key was 0.165 m. The load was introduced at ¼

point over the entire width of 3.0 m. The bridge collapsed by a four-hinge

mechanism at a load of 228 kN. The masonry compression strength was

assumed to be 4 MPa. Cavicchi and Gambarotta (2004) computed the load

capacity of the pure arch without backfill in the range of 46 kN. This

would correspond to 20% of the overall load-bearing capacity. Another

example by Cavicchi and Gambarotta (2004) also gave a load contribution

by the pure arch between 33 and 50% of the overall load-bearing capacity.

See also Cavicchi and Gambarotta (2005).

Royles and Hendry (1991) have also investigated the influence of the

backfill, masonry backup, and spandrel walls on the overall load-bearing

capacity of arch bridges. They reach an improvement compared to the pure

arch of factor 2–12, which would correspond to 8–50%. Becke (2005) has

given an increase of the load by backfill compared to the pure arch of fac-

tor 3 (33%). However, Becke distinguishes between effective and nonef-

fective backfills. This distinction depends on the stiffness of the backfill: If

the stiffness of the backfill is less than 1/100 of the arch, the backfill is not

effective and cannot be considered.

If the backfill is not continuous over the arch, it still can contribute to

the load bearing. Braune (1980) mentions that even transverse and vaults

support the arch.

Latest works about the influence of the backfill and the spandrel walls

are Cavicchi and Gambarotta (2007) and Harvey et al. (2007b).

Besides the backfill, the spandrel walls can also significantly contribute

to the load bearing. For example, Voigtländer (1971) has given a decrease

of stresses of the arch, if the spandrel wall is functioning, of about 20%.

However, this decrease is not found in all areas of the arch over the arch

3.3 Beam Models 131

ing, not at the crown. At the crown, the spandrel walls usually do not reach

good bond conditions. Triebecker has developed a method for the consid-

eration of the spandrel wall in beam models: he simply considers the arch

as fixed beam and supports this beam with an additional beam. This sec-

ond beam should consider the positive effects of the spandrel wall.

Schreyer (1960) also has mentioned the active functioning of the spandrel

walls.

The description of the load-bearing effects of further structural elements

besides the arch itself is already an indicator for the difficulty of the

development of a beam model for arch bridges. In many cases, the models

are heavily disputed. Some codes give general indications such as the Ri

805, which simply states that every model that yields to an equilibrium can

be applied. On the one hand, sometimes models are claimed such as

shown in Fig. 3-15. Here, the limitations of developed hinges during

collapse are considered in that way that under all conditions a moment

has to be applied.

On the other hand, if hinges develop, then the stresses decrease. Model

(1977) suggests a 30% drop of the stresses in the arch after cracks were

initiated. As a consequence, complete hinges could not develop. However,

Model (1977) also considers rather large cracked areas that would corre-

spond with smeared crack models in FE models.

masonry arch bridges under ultimate loads do not show large cracked ar-

eas, but do show great single cracks in regions of maximum loading. In

this region, usually much higher rotations occur than linear-elastic mod-

els show. Figure 3-16 shows test results from Jagfeld and Barthel (2004).

Brencich and Gambarotta (2005) also found such hinges on a bridge over

the Scrivia in Alessandria/Italy.

132 3 Computation of Historical Arch Bridges

(2004)

beam models even if no hinges have been designed and built. Figure 3-17

visualizes this concept. The plasticity theory application for masonry

arches was originally introduced by Kooharian (1952) and Heyman

(1966). An excellent overview of the development of arch models can be

found in Gilbert (2007).

Both methods, the search of an equilibrium inside the arch as static

modelling and the consideration of the kinematic chains as kinematic crite-

ria, are used for the load evaluation process of arch bridges. They can be

seen as upper and lower bound solutions (Fig. 3-18).

Fig. 3-17. Application of plasticity theory for masonry arch bridges (Heyman

1966, Kooharian 1952)

3.3 Beam Models 133

Fig. 3-18. Relationship between upper- and lower-bound solutions (Gilbert 2007)

The MAFEA code in the UIC codes is using nonlinear beam elements,

permitting the development of hinges. Such elements have also been used

in the program Sofistik by Bothe et al. (2004), Herrbruch et al. (2005), and

Mildner (1996). The program STATRA has been used by Möller et al.

(2002). Highly sophisticated programs for the computation of stone arch

bridges are RING (LimitState Ltd 2008) and Archie-M (Harvey 2008).

sections can be found in Hannawald (2006). In general, for the considera-

tion of compound cross sections, a distinction between solid joints and

sliding joints has to be undertaken. If relative sliding inside joints is prohi-

bited, this is called a solid joint. The properties of this compound cross

section are based on geometrical and stiffness properties of single cross

sections. For example, this model can be used for arch bridges by simply

increasing the arch cross section by a certain factor for the partial load

transfer into the backfill. The drawback of this simple model is the fact

that the thickness of the backfill may vary in the length of the arch and

therefore the factor has to be altered over the arch length.

In the following paragraphs, the computation of such a factor will be

shown for solid joints in terms of an effective cross section. Such formulae

have found wide application in structural wood design.

The position a of the line of thrust in such a compound cross section

with solid functions can be computed as

134 3 Computation of Historical Arch Bridges

1 ( EA)2 ⋅ ( h1 + h2 ) (3-58)

a= ⋅ ,

2 ∑ ( EA)i

where E1 and E2 are Young’s modulus for the arch and the backfill, A is

the cross-sectional area, h are the corresponding heights, and the variable a

is shown in Fig. 3-19. If the arch and the backfill have the same height,

then the formula becomes

E2 (3-59)

a = h1 ⋅ .

E1 + E2

( EI )eff = ∑ (( EI )i + ( EA)i ⋅ a2 ) (3-60)

b ⋅ h13 (3-61)

( EI )eff = ( E1 + E2 ) ⋅ + b ⋅ h1 ⋅ ( E1 ⋅ h12 + E2 ⋅ a 2 ) ,

12

where I is the bending stiffness of the single cross section and b is the

width according to Fig. 3-19.

However, if sliding occurs in the joint, the contribution of the backfill to

the load bearing will decrease compared to the solid joint. Such a sliding is

more probable than a solid joint and depends on many factors, such as the

shape of the extrados and material properties of the backfill. The sliding

can be covered in a simplified way by the consideration of a sliding

factor γ. This factor can be computed according to

3.3 Beam Models 135

1 (3-62)

γ=

1 + k1

π ⋅ ( EA)2 ⋅ s1 (3-63)

k1 =

K1 ⋅ l 2

with (EA)2 as the product of Young’s modulus and the cross section size

for cross section 2, which is connected to cross section 1

• K1/s1 as stiffness of the joint between cross sections 1 and 2

• s1 as distance of dowels between cross sections 1 and 2

• l as distance of the moment zero point (l is the span for single-span

beams, 0,8 × li for continuous beams, 2 × lk for cantilevers)

The position of the line of thrust by the same height for the arch and the

backfill and the consideration of the sliding factor becomes

γ ⋅ E2 (3-64)

a = h1 ⋅ .

E1 + γ ⋅ E2

The effective bending stiffness also depends on the sliding factor:

b ⋅ h13 (3-65)

( EI )eff = ( E1 + E2 ) ⋅ + b ⋅ h1 ⋅ ( E1 ⋅ a 2 + γ ⋅ E2 ⋅ ( h2 − a )2 ) .

12

The effective normal force stiffness is computed as

( EA)eff = E1 ⋅ A1 + γ ⋅ E 2 ⋅ A2 . (3-66)

The line of thrust is changing over the length of the arch due to the con-

sideration of the backfill (Fig. 3-20). It can even come out of the arch.

These results fit very well with results by Gocht (1978).

136 3 Computation of Historical Arch Bridges

Becke (2005)

( EA)1 ( EA)1 ⋅ a E1 ( EA)1 ⋅ a (3-67)

N1 = N ⋅ +γ ⋅M⋅ =N⋅ +γ ⋅M⋅

∑ ( EA)i ( EI )eff E1 + E2 ( EI )eff

( EI )1 (3-68)

M1 = M ⋅ .

( EI )eff

The effects of the backfill on the load-bearing behaviour of arches by

using beam models can be done either by applying springs and additional

beams for the backfill or by the introduction of compound cross sections as

shown here. A further model is the application of the finite element

method. This technique not only permits the consideration of the backfill

and possible sliding inside the joints but also allows a general considera-

tion of material and geometry nonlinear behaviour. Therefore, in contrast

to linear-elastic beam models, wherein hinges chosen in advance represent ar-

eas of nonlinear behaviour, this will be done automatically in finite ele-

ment models.

Besides the application of simple beam elements for the modelling of arch

bridges, finite element models of masonry and concrete arch bridges have

become more and more popular. The first finite element analysis of arch

bridges was probably carried out by Towler (1985) and Crisfield (1985).

3.4 Finite Element Method (FEM) 137

grams include modules for realistic material description of masonry and

are used for the simulation of arch bridges. Choo et al. (1991), Dialer

(1991), Loo (1995), Mojsilović (1995), Lourenço (1996), Baker (1997),

Seim and Schweizerhof (1997), Seim (1998), Parikh and Patwardhan

(1999), Huster (2000), Schlegel and Rautenstrauch (2000), Schlegel et al.

(2003), Schlegel (2004), Schlegel and Will (2007), and Schlegel (2008)

have supported this development by developing such realistic material

models for FE programs. Others have simply used already included mate-

rial description modules (Drucker-Prager, Willam Warnke 1974). Such

modules in ANSYS has been used, for example, by Weigert (1996), Wed-

ler (1997), Jagfeld (1998), Melbourne and Tao (1998), Trautz (1998),

Frunzio and Monaco (1998), Fanning and Boothby (2001), Fanning et al.

(2001), Hänel and Reintjes (2001), Mildner and Mildner (2001), Reintjes

(2002), Droese and Bodendiek (2002), Proske (2003), Brencich et al.

(2002), Purtak (2001, 2004), Witzany and Jäger (2005), Purtak et al. (2007),

and Bién et al. (2008). The program ATENA has been used by Schuere-

mans (2001), Cervenka (2004), and Slowik et al. (2005). Healey and

Counsell (1998), Brencich and Colla (2002), Brencich and de Francesco

(2004), and LUSAS (2005) have used the LUSAS program. Stavrouli and

Stavroulakis (2003) and Mura et al. (2003) used the program MARC for the

investigation of historical arch bridges. Rots and van Zijl (2005) have used

DIANA. The program SAP 2000 was used by Toker and Ünay (2004) and

Prader et al. (2008) and the program Strand7 FE analysis was employed by

Ford et al. (2003).

Further examples of finite element models of arch bridges can be found

in Aoki et al. (2004), Aoki and Sabia (2003), Aoki and Sato (2003),

Orlando et al. (2003), Cavicchi and Gambarotta (2006), Drosopoulos et al.

(2007), and Knoblauch et al. (2008). An extension of finite element mod-

els is mentioned by Fuhlrott (2004). The requirement of extensive model-

ling of the foundation and the ground is required by Rombach (2007). And

finally, some authors see general limitations in the application of finite

element models for masonry, such as Chiostrini et al. (1989) and Dialer

(2002).

In general, masonry requires material nonlinear modelling and therefore

finite element models have to adapt to this (Figs. 3-21 to 3-24).

138 3 Computation of Historical Arch Bridges

Fig. 3-21. Failure surface of masonry based on tests according to Page (1983)

Fig. 3-22. Failure surfaces of masonry according to Seim (1998) and Lourenço

(1996)

3.4 Finite Element Method (FEM) 139

Fig. 3-24. Stress-strain behaviour of masonry for tension, shear and compression

according to Schlegel (2004)

(2002), including DEM. However, and that is quite important, the more

complicated the simulation techniques become, the higher a numerical

safety factor gets, as Table 3-19 shows. This increased safety factor re-

flects the increasing uncertainty of an increasing number of input vari-

ables.

Table 3-19. Safety factors for different computation strategies (Lourenço 2002)

Approach/analysis type Safety factor

Allowable stress (fta = 0.2 MPa) 0.31

Kinematic limit analysis 1.8

Geometric safety factor 1.2

Physical nonlinear and no tensile strength 1.8

Physical and geometrical nonlinear and no tensile strength 1.7

Physical nonlinear and tensile strength of 0.2 MPa 2.5

Physical and geometrical nonlinear and tensile strength of 0.2 MPa 2.5

140 3 Computation of Historical Arch Bridges

under great cracks. Then, the advantage of finite element method by

assuming homogenous material properties over certain space regions

cannot hold anymore. An alternative is the application of the discrete

element method (DEM)—sometimes called distinct element method.

In general, the DEM is a procedure for the simulation of movements of

a limited number of bodies with any shape subject to certain interactions.

Single bodies can freely move in space, however, contacts between bodies

are considered. Both static and kinematic boundary conditions are fulfilled.

The static boundaries consider the following in detail:

• Material behaviour functions

• Contact behaviour functions

• External field conditions (Neuberg 2002)

The material behaviour functions describe the behaviour of the single

bodies under external forces. The contract behaviour functions describe the

behaviour of single bodies during interaction and external fields describe

forces such as gravitation that act on all single bodies.

Cinematic conditions are the movement equations of Newton’s second

law and the extension to rotation. Furthermore, for quasistatic solutions,

numerical damping is required.

The overall computation includes the alternate computation of the static

and cinematic relationships over a discrete time. The discrete time also al-

lows the change of the external field forces if required (Neuberg 2002).

Early applications of DEM were done by Cundall (1971) and Cundall

and Strack (1979). Since then, DEM has found wide application in the com-

putation of masonry elements and masonry arch bridges. For example,

Chiostrini et al. (1989), Mamaghani et al. (1999), Bićanić et al. (2002),

Dialer (2002), and Keip and Konietzky (2005) and Lemos (2006) have

used DEM for masonry structures in general.

Maunder (1993), Lemos (1995), Owen et al. (1998), Roberti and Calvetti

(1998), Thavalingam et al. (2001), Brookes and Collings (2003), Bićanić

et al. (2003), Jackson (2004), Schlegel (2004) and Rouxinol et al. (2007)

have used DEM for historical masonry arch bridge computation.

Although DEM is a very general and robust method, the problem for

practical application is still an extensive computation time and a great mul-

titude of different material parameters that are often unknown or difficult

to measure on the structure.

3.6 Comparison of Testing and Modelling 141

data—in this case, the comparison with the investigated ultimate load-

bearing behaviour of real arches. Therefore, for more than a century, tests

on arch bridges were carried out.

First test reports were published by the Austrian engineer and architec-

ture association in 1895. The testing of reinforced and non-reinforced con-

crete arches with a maximum span of 23 m has been described (Huerta

2001). In Germany, the first tests were carried out in 1908 on a three-hinge

road vault with tamped concrete. The bridge was originally constructed in

1902 for an industry exhibition in Düsseldorf. The bridge was designed for

2

a streamroller with 230 kN, and for a crown with 4 kN/m . The vault

reached a span of 28 m. The rise was 2 m. The thickness of the arch

reached 0.75 m at the springing, 0.85 m in the quarter points, and 0.65 m at

the crown. The backfill of the arch was removed. Unfortunately, the

maximum load of the test equipment of 4,230 kN was reached without

failure. Therefore, the maximum load was applied for two days continu-

ously; however, the bridge did not fail. The ratio between observed and

computed ultimate load was more than 18.

In 1935, tests on arch bridges were carried out in the United States

(Jäger 1935). In 1936, Pippard and colleagues carried out tests on arch

bridge models. They discovered that arches develop hinges during failure

and therefore simple linear elastic computations are insufficient for ulti-

mate load-bearing assessment since they do not consider the development

of such hinges. Pippard and Chitty therefore developed a technique called

“instability analysis.” This method showed great affinity with the plastic

methods introduced by Heyman (Das 1995).

At the end of the 1980s and the beginning of the 1990s, about ten arch

bridges were tested in the United Kingdom (eight stone arch bridges and

two models) (Das 1995, BE 16/97). Some of the tests were carried out in

the Bolton Laboratory. These tests, and additional tests ordered by the

Transport and Road Research Laboratory (TRRL, later TRL), resulted in

considerable amount of material for recomputation of the ultimate load of

the bridges. This is especially true for the tests on real arch bridges.

Most of the bridges collapsed by the development of four-hinge mecha-

nisms. Only a minority failed by buckling or exceeding the masonry

142 3 Computation of Historical Arch Bridges

is shown in Table 3-20.

In connection with the tests, the investigation by Royles and Hendry

(1991) should be mentioned. They have classified the tested arches as

follows:

• Compression arch alone

• Compression arch with backfill

• Compression arch with backfill and backup masonry

• Compression arch with backfill, backup masonry and spandrel walls

The tests clearly showed some positive effects of additional structural

elements. For example, load distribution caused by the backfill increased

the load bearing and also backup masonry, stiffer backfill, and stiffer

spandrel walls increased the load capacity of the arches between factor 2

and factor 12 in comparison to the arch only (Royles and Hendry 1991).

Table 3-20. Parameters of the tested bridges (BE 16 1997, RING 2005)

Bridge Span in m Rise in m Arch thickness Width in m Skewness in

in mm degrees

Bridgemil 18.30 2.85 711 8.3 0

Bargower 10.36 5.18 588 8.68 16

Preston 5.18 1.64 360 8.7 17

Prestwood 6.55 1.43 220 3.6 0

Torksey 4.90 1.15 343 7.8 0

Shinafoot 6.16 1.19 542 7.03 0

Strathmashie 9.42 2.99 600 5.81 0

Barlae 9.86 1.69 450 9.8 29

Bundee 4.00 2.00 250 6.0 0

Bolton 6.00 1.00 220 6.0 0

capacity of a single track natural stone arch bridge with tamped concrete.

The tests are intensively presented in Weber (1999) and are also mentioned

in Hoch and Schmitt (1999). The bridge was constructed in 1919 with a

span of 14.0 m and a pier height of 4.20 m. The width of the vault reached

3.65 m. The spandrel walls reached a height of 0.83 m. The arch shape fol-

lowed a basket arch with either three or five circles. The railway line was

closed in 1976 (Weber 1999).

A single load was applied on the railway slab at a quarter point using a

reinforced concrete traverse. The bridge reached an ultimate load of

3,840 kN. At that load, the test had to be stopped since the ultimate load

of the ground anchors used for the anchoring of the concrete traverse had

reached the limit. However, the results showed that the arch overleaped

3.6 Comparison of Testing and Modelling 143

hinge mechanism. The arch showed a horizontal deformation of 4.7 mm

and a vertical deformation of 9.0 mm at the load introduction point. At the

unloaded quarter point, the arch moved upwards about 0.1 mm and hori-

zontally by 3.0 mm. It should be mentioned that at up to 30% of the ulti-

mate load the spandrel walls continued to function together with the arch

and only afterwards partly separated. The rail track also contributed to the

load resistance with a normal force of about 140 kN (Weber 1999).

After the first test, the bridge was partly disassembled: the rail track and

the backfill were cleared. Until a load of 3,450 kN, the development of a

four-hinge mechanism was completed. However, then the load was re-

duced again and the arch returned to the original shape. All failure joints

closed. A further reloading of up to 2,398 kN caused the collapse of the

bridge (Weber 1999).

Since then, many tests have been carried out and the following example

should not be understood as a complete list. Roca and Molins (2004) pre-

sented tests on two bridges with a span of 3.2 m and a rise of 0.65 and

1.6 m, respectively. The bridges were loaded in the quarter points. Details

about the bridges are given in Table 3-21. The bridges failed under 60 and

95 kN, respectively, due to the development of a four-hinge mechanism.

Two hinges already developed at early loading states—one directly under

the load and the other one at the close springing. At 60–75% of the ulti-

mate load, the spandrel wall separated from the arch. The third hinge de-

veloped at 80–90% of the ultimate load, and finally the development of the

fourth hinge at the crown yielded to the collapse. The arch with the lower

rise showed a maximum vertical deformation at the crown of 4.5 cm and

the second arch reached about 2.0 cm.

Vermeltfoort (2001) reports about tests on a masonry arch bridge with-

out backfill in the Netherlands. In the last years Purtak et al. (2007) have

carried out tests at the University of Technology in Dresden, Germany.

Table 3-21. Properties of the bridges tested by Roca and Molins (2004)

Property Bridge c Bridge d

Span 3.20 m 3.20 m

Rise 0.65 m 1.60 m

Overall length 5.20 m 5.20 m

Overall height 0.95 m 2.15 m

Width 1.00 m 1.00 m

Arch thickness 0.14 m 0.14 m

Thickness spandrel wall 0.14 m 0.14 m

Height of backfill at crown 0.10 m 0.10 m

Maximum height of backfill 0.78 m 1.17 m

144 3 Computation of Historical Arch Bridges

Number of tendons for horizontal restrains 6 8

Loading at ¼ Point ¼ Point

Compression strength of stones (longitudinal) 56.8 MPa 56.8 MPa

Young’s modulus of the stone (longitudinal) 12.750 MPa 12.750 MPa

Compression strength of stones (transversal) 51.0 MPa 51.0 MPa

Young’s modulus of the stone (transversal) 10.450 MPa 10.450 MPa

Mortar compression strength (40 × 40 × 80 mm) 8.34 MPa 8.34 MPa

Mortar flexural bending strength (40 × 40 × 160 mm) 2.68 MPa 2.68 MPa

Young’s modulus of the mortar 780 MPa 780 Mpa

Friction coefficient of joint arch backfill 0.33–0.36 0.33–0.36

Compression strength of masonry 21.0 14.0

Density backfill 18 kN/m3 18 kN/m3

computer models of arch bridges. Such comparisons are visualized in

Fig. 3-25 and Table 3-22. However, under normal conditions, the engineer

lacks data of real tests. Therefore, usually other programs are used to esti-

mate the performance of our programs or models. For example, the FILEV

method introduced earlier was tested with results from the program RING.

The results were quite impressive, as shown in Table 3-23 and Fig. 3-26.

3.6 Comparison of Testing and Modelling 145

tigations (B 16 1997, RING 2005)

146 3 Computation of Historical Arch Bridges

investigations (B 16 1997, RING 2005)

Experimental Computed ultimate load using the program

ultimate load CTAP ARCHIE MINIPOINT ARCH MAFEA

Bridgemill 310 183 278 245 217 219

Bargower 560 601 336 350 411 403

Preston 210 184 130 181 73 95

Prestwood 22 0 2 7 6 8

Torksey 108 103 91 124 69 91

Shinafoot 250 268 204 295 205 257

Dundee 104 90 23 123 67 96

Bolton 117 41 39 124 43 52

Strathmashie 132 118 142 112 109 120

Barlae 290 232 216 320 182 165

FILEV and RING software (Martín-Caro and Martínez 2004)

ing FILEV and RING software (Martín-Caro and Martínez 2004)

Arch Ultimate load Ultimate load Ratio

Span Rise thickness Coverage at according to according to

Nr. in m in m in m crown in m FILEV in kN RING in kN

1 5.0 2.50 0.500 0.5 1,541 1,525 1.0105

2 5.0 1.25 0.500 0.5 1,726 2,135 0.8084

3 7.5 3.75 0.750 0.5 2,376 2,438 0.9746

4 7.5 1.88 0.750 0.5 2,662 3,175 0.8384

5 7.5 3.75 0.675 0.5 1,456 1,580 0.9215

3.6 Comparison of Testing and Modelling 147

Span Rise thickness Coverage at according to according to

Nr. in m in m in m crown in m FILEV in kN RING in kN

6 7.5 1.88 0.675 0.5 1,630 2,010 0.8109

7 10.0 5.00 0.700 0.5 1,165 1,393 0.8363

8 12.5 2.50 0.700 0.5 1,305 1,400 0.9321

9 12.5 6.25 0.750 0.5 1,312 1,405 0.9338

10 12.5 2.50 0.750 0.5 1,666 1,790 0.9307

11 12.5 6.25 0.875 0.5 1,841 1,798 1.0239

12 12.5 2.50 0.875 0.5 2,338 2,640 0.8856

13 15.0 7.50 0.900 0.5 1,928 1,890 1.0201

14 15.0 3.00 0.900 0.5 2,449 2,390 1.0247

15 18.0 9.00 0.900 0.5 1,750 1,750 1.0000

16 18.0 3.60 0.900 0.5 2,046 2,000 1.0230

17 20.0 10.00 1.000 0.5 1,887 1,790 1.0542

18 20.0 4.00 1.000 0.5 2,397 2,475 0.9685

sity of Dundee. It heavily relates to works by Heyman. However, addition-

ally earth pressure can be included in the computation. The program first

investigates the location of the hinges in the arch under live and dead load.

Then the line of thrust is computed. ARCHIE then gives a lower limit of

the load-bearing capacity. The program did not give deformations, multi-

layered arches could not be considered, the failure criterion was always the

development of a four-hinge mechanism, multi-span arches could not be

investigated, and concrete backfill and vaults also could not be modelled

(Brookes and Collings 2003).

In 1999, the program ARCHIE-M was presented by Obvis Ltd in the

United Kingdom. The program first showed only slight differences from the

ARCHIE, such as the consideration of ground pressure and the distribution

of single loads in longitudinal direction (Brookes and Collings 2003). How-

ever, since then, the program has been permanently extended and updated. A

demo version can be downloaded at Harvey (2008) or at Obvis Ltd (2006).

The program RING was originally developed during the 1990s at the

Bolton Institute and the University of Sheffield. In 2007, the second

edition was released. It considers the development of a mechanism, but

shear failure can also be considered. Furthermore, some special conditions

can be considered, such as multi-ring arches. The original version could

not consider concrete backfill and vaults. Furthermore, no deformation was

given and buckling was not considered (Brookes and Collings 2003). A

trial version of RING 1.5 can be downloaded at the University of Sheffield

(2008), and a demo version of RING 2.0 at LimitState Ltd (2008).

The Italian program SAV (Stabilita`di Archie Volte in Matura) has been

introduced by AEDES (2008). Another Italian program is called ARCO

148 3 Computation of Historical Arch Bridges

Gelfi (2006). Smars (2006) has developed the program Calipous for the

computation of gothic vaults and structures of masonry arches in Belgium.

The German railway often uses the program GETRA for the load

estimation of historical railway arch bridges (Ludwig 2000), and the Saxon

road departments often use Excel sheets developed by Dibeh et al. (1997)

based on linear-elastic arch computation using the theory of Strassner. Also,

the consultant company Pauser has developed a program (Heinlein 2008,

Pauser 2007). Teaching programs for students regarding the mechanisms of

arches were developed by Greenwold et al. (2008) during the Active Statics

project and by Block (2006).

The presented beam models have not considered the transverse action of

loads inside the arch bridges: they simply considered only longitudinal ef-

fects. However, the consideration of transversal effects may decrease the

load on the arch and offer an opportunity to permit higher traffic loads on

arch bridges. Therefore, the load spreading in transverse direction is usu-

ally considered in terms of effective width, as shown in Fig. 3-27.

Fig. 3-27. Effective width based on the height of backfill and height of the arch

The spreading angle of the load can be taken from several codes of prac-

tice or recommendations. For example,

• For gravel or sand, an angle of 30° is used according to the DIN 1055-1

• For concrete or masonry, 45° is assumed by Haser and Kaschner (1994)

• For masonry, about 30° is assumed according to DIN 1052-1 (1996)

3.7 Transverse Direction (Effective Width) 149

Whereas the current German codes (DIN 1075 or DIN-report 101) for road

bridges lack information about the effective width, former codes included

some recommendations, such as

• TGL 12999 (3/1977) “Re-computation of existing bridges”

• SBA-regulation 169/89 (9/1989) “Bridges for traffic – recomputation of

road traffic bridge built of concrete and masonry”

Based on these former recommendations Haser and Kaschner (1994) give,

the following rules:

One-lane road Two-lane road

bm = 4 m bm = 7 m

bm = 0.25× l

bm ≤ b

However, the width must not be wider than the bridge.

For railway bridges, the Ril 804 gives a spreading function. The general

assumption for such a spreading is the coverage height of bridges. Bridges

are considered as covered, if the coverage is greater than 60 cm. Figure

3-28 shows an example of a railway bridge with and without concrete

plate. As shown in Fig. 3-28, the effective width alters with the length of

the arch. For railway bridges, therefore, the Ril 804/805 permits the com-

putation of a middle effective width bm

bc + bs (3-69)

bm =

2

with bc as the effective width at crown and bs as the effective width at

springing.

However, if the force spreading exceeds the bridge width, the force cre-

ates a horizontal loading on the spandrel walls as shown in Fig. 3-29.

150 3 Computation of Historical Arch Bridges

Fig. 3-28. Example of load spreading in an arch bridge. In the top subfigure, the

bridge is shown with a railway plate, and in the subfigure in the middle and at the

bottom, the bridge is in the original state at the crown and at the springing

3.7 Transverse Direction (Effective Width) 151

Fig. 3-29. If the load spreading reaches the spandrel wall, the load can cause hori-

zontal pressure on the spandrel wall as shown in this example

models can be used such as the models by Harvey (2006) or by the use of

FEM as shown by Nautiyal (1993) and Frenzel (2004). Figure 3-30 shows

the different load-spreading models by Harvey, whereas Fig. 3-31 shows

the results from a FEM computation in terms of vertical stresses inside an

arch bridge caused by traffic load.

Fig. 3-30. Force flow models in longitudinal and transversal direction according to

Harvey (2006)

152 3 Computation of Historical Arch Bridges

Fig. 3-31. Vertical stress at the ¼ point of the arch according to Frenzel (2004)

References

Aita D, Foce F, Barsotti R & Bennati S (2007) Collapse of masonry arches in

Romanesque and Gothic constructions. Proceedings of the Arch`07 – 5th In-

ternational Conference on Arch Bridges, PB Lourenco, DV Oliveira & A Por-

tela (Eds), 12–14 September 2007, Madeira, pp 625–632

Aoki T & Sabia D (2003) Collapse Analysis of Masonry Arch Bridges, Proceed-

ings of the 9th International Conference on Civil and Structural Engineering

Computing, Egmond aan Zee, The Netherlands, 2–4, September 2003, BHV

Topping (Ed), Civil Comp Press, Stirling, United Kingdom, 2003, pp 1–14

(CD-ROM)

Aoki T & Sato T (2003) Application of Bott-Duffin Inverse to Static and Dynamic

Analysis of Masonry Structures, Structural Studies, Repairs and Maintenance

of Heritage Architecture VIII, pp 277–286

Aoki T, Komiyama D, Sabia D & Rivella R (2004) Theoretical and Experimental

Dynamic Analysis of Rakanji Stone Arch Bridge, Honyabakei, Oita, Japan,

Proceedings of 7th International Conference on Motion and Vibration Control

MOvIC’04, St. Louis, 8.–11. August, 2004, pp 1–9 (CD-ROM)

Audenaert A, Peremans H & Reniers G (2007) Journal of Engineering Mathemat-

ics, 59, pp 323–336

B 16 (1997) Comparison of Masonry Arch Bridge Assessment Methods. Annex E,

Volume 3, Section 4, Part 4, http://www.archive2.official-documents.co.uk/

document/deps/ha/dmrb/vol3/sect4/ba1697h.pdf, 2005

Baker M (1997) FE analysis of Masonry and Application to masonry Arch

Bridges. PhD Thesis, Cardiff University

BE 16/97 (2005) Comparison of masonry arch bridge assessment methods. Annex

E, Volume 3, Section 4, Part 4, BE 16/97, www.archive2.official-documents.

co.uk/document/deps/ha/dmrb/vol3/sect4/ba1697h.pdf

References 153

Berücksichtigung der Mitwirkung der Hinterfüllung bei historischen

Natursteinbogenbrücken. Diplomarbeit Technische Universität Dresden

Bićanić N, Stirling C & Pearce CJ (2002) Discontinuous Modelling of Structural

Masonry. Fifth World Congress on Computational Mechanics, HA Mang, FG

Rammerstorfer & J Eberhardsteiner (Eds), July 7–12 2002, Vienna, Austria

Bićanić N, Stirling C & Pearce CJ (2003) Discontinuous modelling of masonry

bridges. Computational Mechanics, 31 (1–2), May 2003, pp 60–68

Bién J, Kamiński T & Trela C (2008) Numerical analysis of old masonry bridges

supported by field tests. Bridge Maintenance, Safety, Management, Health

Monitoring and Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis

Group, London, pp 2384–2391

Bienert G (1959–60) Aufgaben der Modellstatik bei der Konstruktion und

Unterhaltung der Verkehrsbauwerke. Wissenschaftliche Zeitschrift der

Hochschule für Verkehrswesen „Friedrich List“ Dresden, 1959/60, Heft 1

Bienert G, Sauer & Schmidt (1960–62) Beitrag zur Ermittlung von Tragreserven

in massiven Brückengewölben. Wissenschaftliche Zeitschrift der Hochschule

für Verkehrswesen “Friedrich List” Dresden, 1960/61, Heft 2 und 9, 1961/62,

Heft 1

Block P (2006) InteractiveTHRUST.

http://web.mit.edu/masonry/interactiveThrust/examples.html

Bothe E, Henning J, Curbach M, Bösche T & Proske D (2004) Nichtlineare

Berechnung alter Bogenbrücken auf der Grundlage neuer Vorschriften. Beton-

und Stahlbetonbau 99, Heft 4, pp 289–294

Braune W (1980) Über die Tragwirkung von Gewölbebrücken mit

längsgegliederten massiven Aufbauten. Die Bautechnik, 57 (2), pp 37–45 and

(3), pp 93–100

Brencich A & Colla C (2002) The influence of construction technology on the

mechanics of masonry railway bridges, Railway Engineering 2002, 5th Inter-

national Conference, 3–4 July 2002, London

Brencich A & de Francesco U (2004) Assessment of Multispan Masonry Arch

Bridges: Simplified Approach. ASCE Journal of Bridge Engineering, Novem-

ber/December 2004, pp 582–590

Brencich A & Gambarotta L (2005) Mechanical response of solid clay brickwork

under eccentric loading. Part I. Unreinforced masonry. Materials and Struc-

tures, 38(2), pp 257–266

Brencich A, Corradi C, Gambarotta G, Mantegazza G & Strerp E (2002) Com-

pressive Strength of Solid Clay Brick Masonry under Eccentric Loading.

Department of Structural and Geotechnical Engineering, University of

Genoa, Italy

Brencich A, De Francesco U, Gambarotta L (2001) Elastic no tensile resistant –

plastic analysis of masonry arch bridges as an extension of castigliano’s

method. 9th Canadian masonry symposium

Brookes C & Collings M (2003) ARCHTEC – Verification of Structural Analysis.

Gifford and Partners Document No: B1660A/V10/R03

Brühwiler E (2008) Rational approach for the management of medium size bridge

stock. Bridge Maintenance, Safety, Management, Health Monitoring and In-

154 3 Computation of Historical Arch Bridges

formatics, H Koh & DM Frangopol (Eds), Taylor & Francis Group, London,

pp 2642–2649

Busch P & Zumpe G (1995) Tragfähigkeit, Tragsicherheit und Tragreserven von

Bogenbrücken. 5. Dresdner Brückenbausymposium, Fakultät Bauingenieurwesen,

Technische Universität Dresden, pp 153–169

Cavicchi A & Gambarotta L (2004) Upper bound limit analysis of multispan ma-

sonry bridges including arch-fill interaction. Arch Bridges IV – Advances in

Assessment, Structural Design and Construction, P Roca & C Molins (Eds),

CIMNE, Barcelona, pp 302–311

Cavicchi A & Gambarotta L (2005) Collapse analysis of masonry bridges taking

into account arch-fill interaction. Engineering Structures, 27, pp 605–615

Cavicchi A & Gambarotta L (2006) Two-dimensional finite element upper bound

limit analysis of masonry bridges. Computers & Structures, 84, pp 2316–2328

Cavicchi A & Gambarotta L (2007) Load carrying capacity of masonry bridges:

numerical evaluation of the influence of fill and spandrels. Proceedings of the

Arch`07 – 5th International Conference on Arch Bridges, PB Lourenco, DV

Oliveira & A Portela (Eds), 12–14 September 2007, Madeira, pp 609–616

Cervenka V (2004) Cervenka-Consulting: http://www.cervenka.cz

Chiostrini S, Foraboschi P & Sorace S (1989) Problems connected with the ar-

rangement of a non-linear finite element method to the analysis of masonry

structures. Structural Repair and Maintenance of Historical Buildings. Com-

putational Mechanics Publications Birkhäuser

Choo BS, Coutie MG & Gong NG (1991) Finite-element analysis of masonry arch

bridges using tapered elements. Proceedings of the Institution of Civil Engi-

neers, Part 2, December 1991, pp 755–770

Corradi M (1998) Empirical methods for the construction of masonry arch bridges

in the 19th century. Arch Bridges: history, analysis, assessment, maintenance

and repair. Proceedings of the second international arch bridge conference,

Edr A. Sinopoli, Venice 6–9 October 1998, Balkema Rotterdam, pp 25–36

Corradi M & Filemio V (2004) A brief comparison between mechanical aspects

and construction of arch bridges during the XVIIIth and XIXth centuries.

Arch Bridges IV – Advances in Assessment, Structural Design and Construc-

tion, P Roca & C Molins (Eds), CIMNE, Barcelona, pp 79–86

COST-345 (2004) European Commission Directorate General Transport and En-

ergy: COST 345 – Procedures Required for the Assessment of Highway

Structures: Numerical Techniques for Safety and Serviceability Assessment –

Report of Working Groups 4 and 5

Craemer H (1943) Die “erzwungene Drucklinie” als Ausdruck der versteifenden

Wirkung der Übermauerung von Gewölben. Beton- und Stahlbetonbau, Heft

9/10

Crisfield M A (1985) Finite Element and Mechanism Methods for the Analysis of

Masonry and Brickwork Arches. Transport and Road Research Laboratory,

London

Cundall PA & Strack ODL (1979) A discrete numerical model for granular as-

semblies. Géotechnique, 29, pp 47–65

References 155

movements in blocky rock systems. In Proceedings of the Symposium of the

International Society of Rock Mechanics. Vol. 1, Paper II-8, Nancy, France

Das PC (1995) The Assessment of Masonry Arch Bridges. The Aesthetics of Load

bearing Masonry Arch Bridges. Arch Bridges, C. Melbourne (Ed), Proceed-

ings of the 1st International Conference on Arch Bridges, Bolton, Thomas

Telford, London, pp 21–27

Dialer C (1991) Bruch- und Verformungsverhalten von schubbeanspruchten

Mauerwerksscheiben. Mauerwerkskalender 1992, Ernst & Sohn, Berlin

Dialer C (2002) Ausgewählte Schäden und Schäden an Instandsetzungen von

Massivbauwerken. Massivbau 2002, Forschung, Entwicklungen und

Anwendungen. Hrsg. Konrad Zilch, Technische Universität München,

pp 218–237

Diamantidis D (2001) Assessment of Existing Structures, Joint Committee of

Structural Safety JCSS, Rilem Publications S.A.R.L.: The publishing Com-

pany of Rilem, Volume 159

Dibeh T, Pfeifer T & Hein S (1997) Statische Berechnung von Bogenbrücken &

Statische Berechnung von Natursteinbogenbrücken. Saxon institute for road

construction, Rochlitz

Dischinger F (1949) Massivbau. Taschenbuch Bauingenieure. F Schleicher (Ed),

Reprint 1949, Springer Verlag Berlin, 1943

DR (1985) Deutsche Reichsbahn: Richtlinie Einbau von Fahrbahnwannen. year

uncertain

Droese S & Bodendiek P (2002) Die Unterfangung des Bahrmühlenviaduktes,

Rechnerischer Nachweis des Viaduktes für Lasten und die Zwangseinwirkung

aus dem Bauvorgang. Die Bautechnik 79, pp 455–463

Drosopoulos GA, Stavroulakis GE & Massalas CV (2007) FRP reinforcement of

stone arch bridges: Unilateral contact models and limit analysis. Composites:

Part B 38, pp 144–151

Enevoldsen I (2008) Practical implementation of probability based assessment

methods for bridges. Bridge Maintenance, Safety, Management, Health Moni-

toring and Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis

Group, London, pp 83–104

Fain JS (1953) Beispiele zur Berechnung von Stein- und Bogenbrücken. VEB

Verlag Technik, Berlin

Falter H (1998) Untersuchung historischer Wölbkonstruktionen – Auswirkungen der

Herstellung und des Baumaterials. Dissertation an der Universität Stuttgart

Fanning PJ & Boothby TE (2001) Three-dimensional modelling and full-scale

testing of stone arch bridges. Computers and Structures, 79, pp 2645–2662

Fanning PJ, Boothby TE & Roberts BJ (2001) Longitudinal and transverse effects

in masonry arch assessment. Construction and Building Materials, 15,

pp 51–60

Fischer U (1940–41) Ausnutzung des Zusammenwirkens von Bogen und Aufbau.

Internationaler Verein für Brücken- und Hochbau, Abhandlung Band 61

Fischer U (1942) Die Mitwirkung des Aufbaus massiver Bogenbrücken. Beton

und Eisen, 37. Jahrgang, Heft 19, pp 310–315

156 3 Computation of Historical Arch Bridges

Ford TE, Augarde CE & Tuxford SS (2003) Modelling masonry arch bridges us-

ing commercial finite element software, 9th International Conference on Civil

and Structural Engineering Computing, Egmond aan Zee, The Netherlands,

2–4 September 2003

Frenzel M (2004) Untersuchung zur Quertragfähigkeit von Bogenbrücken aus

Natursteinmauerwerk. Diplomarbeit Technische Universität Dresden

Frunzio G & Monaco M (1998) An approach to the structural model for masonry

arch bridges: Pont Saint Martin as a case study. Arch Bridges: History, Analy-

sis, Assessment, Maintenance and Repair. Proceedings of the Second Interna-

tional Arch Bridge Conference, A Sinopoli (Ed), Venice 6.–9. October 1998,

Balkema Rotterdam, pp 231–240

Fuhlrott A (2004) Grundlagen der Extended Finite Elemente Methode (X-FEM).

Studienarbeit am Institut für Strukturmechanik, Fakultät Bauingenieurwesen,

Bauhaus-Universität Weimar

Gelfi P (2006) Arco.exe http://dicata.ing.unibs.it/gelfi/arco.htm

Gilbert M (2007) Limit analysis applied to masonry arch bridges: state-of-the-art

and recent developments. Proceedings of the Arch`07 – 5th International Con-

ference on Arch Bridges. PB Lourenco, DV Oliveira & A Portela (Eds), 12–

14 September 2007, Madeira, pp 13–28

Gilbrin G (1913) Störungen des normalen Zustandes in Brückengewölben.

Dissertation an der damaligen Kgl. TH zu München, Verlag Ernst und Sohn,

Berlin 1913

Gocht R (1978) Untersuchungen zum Tragverhalten rekonstruierter Eisenbahnge-

wölbebrücken, Dissertation Hochschule für Verkehrswesen “Friedrich List”

Dresden

Greenwold S, Allen E & Zalewski W (2008) Active Statics. http://acg.media.

mit.edu/people/simong/statics/Start.html

Gutermann M (2002) Ein Beitrag zur experimentell gestützten Tragsicherheitsbe-

wertung von Massivbrücken, Dissertation, Fakultät Bauingenieurwesen

Haase H (1899) Theorie des parabolischen Brückengewölbes oder das

Grundgesetz des Horizontalschubs in seiner Anwendung auf das

Brückengewölbe unter der ausschließlichen Wirkung vertikaler Aussenkräfte,

entwickelt an dem Beispiel einer gewölbten Bahnüberbrückung. Kommis-

sionsverlag der Nationalen Verlagsanstalt, Regensburg

Hänel D & Reintjes KH (2001) Die Unterfangung des Bahrmühlenviaduktes,

Planung, Entwurf und Vergabe. Die Bautechnik, 78, pp 543–555

Hannawald F (2006) Zur physikalisch nichtlinearen Analyse von Verbund-Stab-

tragwerken unter quasi-statischer Langzeitbeanspruchung. Dissertation,

Technische Universität Dresden, Institut für Stahl- und Holzbau

Harvey B (2006) Some problems with arch bridge assessment and potential

solutions. Structural Engineer, 84 (3), 7th February 2006, pp 45–50

Harvey B (2008) Archie-M. http://billharvey.typepad.com/archi_m/

Harvey B, Ross K & Orbán Z (2007a) A simple, first filter assessment for arch

bridges. Proceedings of the Arch`07 – 5th International Conference on Arch

Bridges, PB Lourenco, DV Oliveira & A Portela (Eds), 12–14 September

2007, Madeira, pp 275–280

References 157

Harvey WJ (2007) Rule of thumb method for the assessment of arches. UIC Re-

port (draft)

Harvey WJ, Maunder EAW & Ramsay ACA (2007b) The influence of the span-

drel wall construction on arch bridge behaviour. Proceedings of the Arch`07 –

5th International Conference on Arch Bridges, PB Lourenco, DV Oliveira &

A Portela (Eds), 12–14 September 2007, Madeira, pp 601–608

Haser H & Kaschner R (1994) Spezielle Probleme bei Brückenbauwerken in den

neuen Bundesländern. Berichte der Bundesanstalt für Straßenwesen, Verlag

für neue Wissenschaft, Bergisch Gladbach

Healey TN & Counsell JHW (1998) Widening and strengthening of Londons

Kingston Bridge. Arch Bridges: history, analysis, assessment, maintenance

and repair. Proceedings of the second international arch bridge conference,

Edr. A. Sinopoli, Venice 6.–9. October 1998, Balkema Rotterdam, pp 389–398

Heinlein M (2008) Pauser Ingenieurbüro. Berechnung von gemauerten

Gewölbebrücken. One-on-one interview on the 16.12.2008

Heinrich B (1983) Brücken – Vom Balken zum Bogen. Rowohlt Taschenbuch

Verlag GmbH, Hamburg

Herrbruch J, Groß JP & Wapenhans W (2005) Wirklichkeitsnähere Tragwerksanalyse

von Gewölbebrücken. Mauerwerk 9, Heft 3, pp 89–92

Herzog R (1962) Wechselbeziehungen zwischen Bogen und Bogenüberbauten von

Brücken. Technisch-Wissenschaftlicher Verlag des Ministeriums für

Automobiltransport und Chaussee-Straßen der RSFSR, Moskau

Heyman J (1966) The stone skeleton. International Journal of Solids and Struc-

tures, 2, pp 249–279

Heyman J (1998) The assessment of strength of masonry arches. Arch Bridges:

History, Analysis, Assessment, Maintenance and Repair. Proceedings of the

Second International Arch Bridge Conference, Edr A Sinopoli, Venice 6.–9.

October 1998, Balkema Rotterdam, pp 95–98

Hoch G & Schmitt B (1999) Nebenbahnen in der Oberpfalz, Verlag Michael

Resch, Coburg

Huerta S (2001) Mechanics of masonry vaults: The equilibrium approach. Histori-

cal Constructions, PB Lourenço & P Roca (Eds), Guimarães, pp 47–69

Huerta S (2004) Arcos, bóvedas y cúpulas. Instituto Juan de Herrera. Escuela

Técnica Superior de Arquitetura Madrid.

Huges TG & Blackler MJ (1997) A review of the UK masonry arch assessment

methods. Proceedings of Institution Civil Engineering Structures and Build-

ings, 122, August, pp 305–315

Huster U (2000) Tragverhalten von einschaligem Natursteinmauerwerk unter

zentrischer Druckbeanspruchung: Entwicklung und Anwendung eines Finite-

Elemente-Programmes. Kassel University Press: Kassel

Jackson P (2004) Highways Agency BD on New Masonary Arch Bridges. 27th

February 2004. http://www.smart.salford.ac.uk/February_2004_Meeting.php

Jäger A (1935) Amerikanische Versuche an den Modellen eines eingespannten

Bogens und einer Bogenreihe. Beton- und Eisen, 34. Jahrgang, Heft 21,

pp 338–344

158 3 Computation of Historical Arch Bridges

Gewölbereihe unter Berücksichtigung der Hintermauerung als mittragende

Masse. Verlag Ernst & Sohn, Berlin

Jagfeld M & Barthel R (2004) Zur Gelenkbildung in historischen Tragsystemen

aus Mauerwerk. Bautechnik (81), Heft 2, pp–96102

Jagfeld M (1998) Tragverhalten gemauerter Wölbkonstruktionen bei großen

Auflagerverschiebungen – Vergleich verschiedener FE Modelle. 16. CAD-

FEM User’s Meeting, 7.–9. Oktober 1998 in Bad Neuenahr

Jensen JS, Plos M, Casas JR, Cremona C, Karoumi R & Melbourne C (2008)

Guideline for load and resistance assessment of existing European railway

bridges. Bridge Maintenance, Safety, Management, Health Monitoring and In-

formatics, H Koh & DM Frangopol (Eds), Taylor & Francis Group, London,

pp 3658–3665

Keip MA & Konietzky H (2005) Mauerwerk unter dynamischen Belastungen –

Untersuchungen mittels der DEM, Mauerwerkstag Essen, 3. Mai 2005

Knoblauch F-J, Laubach A, Sauer W, Schmidt M & Sprinke P (2008) Erneuerung

des Burtscheider Viadukts in Aachen. Beton- und Stahlbetonbau 103, Heft 3,

pp 195–203

Kooharian A (1952) Limit analysis of voussoir (segmental) concrete arches.

Journal American Concerte Institute, 24, pp 317–328

Kurrer K-E (2002) Geschichte der Baustatik. Ernst und Sohn Verlag für

Architektur und technische Wissenschaften GmbH, Berlin 2002

Lachmann H (1990) Über die Standsicherheit gemauerter Gewölbebrücken,

Bautechnik 67, Heft 2, pp 61–63

Leliavsky S (1982) Arches and short span bridges. Design Textbook in Civil En-

gineering: Volume VII, Chapman and Hall, London

Lemos JV (1995) Assessment of the ultimate load of a masonry arch using dis-

crete elements. Computer Methods in Structural Masonry – 3, GN Pande &

J Middleton (Eds), Books Journals International 1995, pp 294–302

Lemos JV (2006) Modelling of historical masonry with discrete elements. Compu-

tational Mechanics – Solids, Structures and Coupled Problems, Springer,

pp 375–391

LimitState Ltd (2008) http://www.masonryarch.com

Loo Y-C (1995) Collapse load analysis of masonry arch bridges. Arch Bridges.

Proceedings of the 1st International Conference on Arch Bridges, C. Melbourne

(Ed), Bolton, Thomas Telford, London, pp 167–174

Lourenço PB (1996) Computational Strategies for Masonry Structures. Dissertation,

Technische Universität Delft, Delft University Press

Lourenço PB (2002) Computations on historic masonry structures, Structure En-

gineering Materials, 4, pp 301–319

Ludwig D (2000) Programm zur Ermittlung der Tragfähigkeit und der

Beförderungsbedingungen sowie zur Einstufungsberechnung an

Gewölbebrücken der Bewertungsstufe 2, DB Netz AG, Magdeburg

LUSAS (2005): Medieval masonry arch passes assessment with LUSAS

www.lusas.com/case/bridge/masonry_arch.html

Mamaghani IHP, Aydan Ö & Kajikawa Y (1999) Analysis of masonry structures

under static and dynamic loading by discrete finite element method. Journal of

References 159

gineers (JSCE), Japan, No. 626/I-48, 16, pp 75–86

Martín-Caro JA & Martínez LJ (2004) A first level structural analysis tool for the

Spanisch Railways Masonry Arch Bridges. Arch Bridges IV – Advances in

Assessment, Structural Design and Construction, P Roca & C Molings (Eds),

CIMNE, Barcelona, pp 192–201

Martinez JA, Revilla J & Aragon Torre A (2001) Critical thickness criteria on

stone arch bridges with low rise/span ratio and current traffic loads. Historical

Constructions, PB Lourenço & P Roca (Eds), Guimarães, pp 609–616

Maunder EAW (1993) Limit analysis of masonry structures based on discrete

elements. Structural Repair and Maintenance of Historical Building III

(STREMAH), CA Brebbia & RJB Frewer (Ed), Computational Mechanics

Publication, Glasgow, pp 367–374

Melbourne C & Tao H (1998) The behaviour of open spandrel brickwork arch

bridges. Arch Bridges: History, Analysis, Assessment, Maintenance and Re-

pair. Proceedings of the Second International Arch Bridge Conference, A Si-

nopoli (Ed), Venice 6.–9. October 1998, Balkema Rotterdam, pp 263–269

Mildner K & Mildner A (2001) Bauwerksmessungen vor und nach der

Instandsetzung einer Eisenbahn-Gewölbebrücke und statische Modellbildung.

Ehrenkolloquium Prof. Dr.-Ing. Möller zum 60. Geburtstag, Lehrstuhl für

Statik und Dynamik der Tragwerke, Technische Universität Dresden, Dresden

Mildner K (1996) Tragfähigkeitsermittlung von Gewölbebrücken auf der

Grundlage von Bauwerksmessungen. 6. Dresdner Brückenbausymposium,

Fakultät Bauingenieurwesen, Technische Universität Dresden, pp 149–162

Model P (1977) Beitrag zur Berücksichtigung der Gewölbehintermauerung, der

Rissbildung und der elastischen Stützung durch den Baugrund und die

Dammschüttung bei der statischen Berechnung einfeldriger Gewölbebrücken,

Dissertation Hochschule für Verkehrswesen “Friedrich List” Dresden

Modena C, Valluzzi MR, da Porto F, Casarin F & Bettio C (2004) Structural up-

grading of a brick masonry arch bridge at the Lido (Venice). Arch Bridges IV –

Advances in Assessment, Structural Design and Construction, P Roca & C

Molins (Eds), CIMNE, Barcelona, pp 435–443

Mojsilović N (1995) Zum Tragverhalten von kombiniert beanspruchten Mauer-

werk. Dissertation, IBK Bericht Nr. 216, ETH Zürich, Birkhäuser Verlag, Ba-

sel

Molins C & Roca P (1998) Capacity of masonry arches and spartial structures.

Journal of Structural Engineering ASCE, 124 (6), pp 653–663

Möller B, Graf W, Beer M & Bartzsch M (2002) Sicherheitsbeurteilung von

Naturstein-Bogenbrücken. 6. Dresdner Baustatik-Seminar. Rekonstruktion

und Revitalisierung aus statisch-konstruktiver Sicht. 18. Oktober 2002.

Lehrstuhl für Statik + Landesvereinigung der Prüfingenieure für Bautechnik

Sachsen + Ingenieurkammer Sachsen. Technische Universität Dresden,

pp 69–91

Mura I, Odoni Z & Perra M (2003) Static analysis of the Turris Lybisonis Roman

multispan stone arch bridge. Structural Studies, Repairs and Maintenance of

Heritage Architecture VIII, CA Brebbia (Ed), Southampton, WIT-Press 2003,

pp 667–674

160 3 Computation of Historical Arch Bridges

mensional Collapse of Masonry Arch. PhD. Cambridge University

Neuberg C (2002) Ein Verfahren zur Berechnung des räumlichen passiven

Erddrucks vor parallel verschobenen Trägern. Institut für Geotechnik,

Technische Universität Dresden, Mitteilungen Heft 11, Dresden

O’Connor A, Pedersen C, Enevoldsen I, Gustavsson L & Hammarbäck J (2008)

Probability Based Assessment of a Large Riveted Truss Railway Bridge.

Bridge Maintenance, Safety, Management, Health Monitoring and Informat-

ics, H Koh & DM Frangopol (Eds), Taylor & Francis Group, London, pp

1812–1819

Obvis Ltd (2006) http://www.obvis.com/

Orlando M, Spinelli P & Vignoli A (2003) Structural analysis for the reconstruc-

tion design of the old bridge of Mostar. Structural Studies, Repairs and Main-

tenance of Heritage Architecture VIII, Edr Brebbia, WIT-Press, pp 617–626

Owen DR, Peric D, Petric N, Brookes CL & James PJ (1998) Finite/discrete ele-

ment models for assessment and repair of masonry structures. Arch Bridges:

History, Analysis, Assessment, Maintenance and Repair. Proceedings of the

Second International Arch Bridge Conference, A Sinopoli (Ed), Venice 6.–9.

October 1998, Balkema Rotterdam, pp 173–180

Page AW (1983) The strength of brick masonry under biaxial tension-

compression. International Journal of Masonry Construction, 3(1), pp 26–31

Parikh SK & Patwardhan YG (1999) Widening of an existing old masonry arch

bridge through RCC box girder deck. Current and Future Trends in Bridge

Design, Construction and Maintenance in Bridge Design Construction and

Maintenance, PC Das, DM Frangopol & AS Nowak (Eds), Thomas Telford,

London, pp 218–226

Pauser A (2005) Ingenieurbüro: Guide to the high-level assessment of masonry

arch bridges. UIC Report (draft)

Prader J, Weidner JW, Hassanain H, Moon F, Aktan E, Jalinoos F, Buchanan B &

Ghasemi H (2008) Load Testing and Analysis of Bridges Missing Critical

Documentation. Bridge Maintenance, Safety, Management, Health Monitor-

ing and Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis Group,

London, pp 1877–1885

Proske D (2003) Ein Beitrag zur Risikobeurteilung alter Brücken unter

Schiffsanprall. Dissertation, Technische Universität Dresden

Purtak F (2001) Tragfähigkeit von schlankem Mauerwerk. Dissertation.

Technische Universität Dresden, Fakultät Architektur

Purtak F (2004) Zur nichtlineare Berechnung von Bogenbrücken aus

Natursteinmauerwerk. 22. CAD-FEM Users Meeting 2004 – International

Congress on FEM Technology with ANSYS CFX & ICEM CFD Conference,

10.–12. November 2004, Dresden

Purtak F, Geißler K & Lieberwirth P (2007) Bewertung bestehender

Natursteinbogenbrücken. Bautechnik 8, Heft 8, pp 525–543

Reintjes K-H (2002) Die Unterfangung des Bahrmühlenviaduktes bei Chemnitz –

Stand des Bauvorhabens. 12. Dresdner Brückenbausymposium – Planung,

Bauausführung und Ertüchtigung von Massivbrücken. 14. März 2002.

References 161

pp 145–163

RING (2005) Technical details. http://www.shef.ac.uk/ring/technical.html

Roberti MG & Calvetti F (1998) Distinct element analysis of stone arches. Arch

Bridges: History, Analysis, Assessment, Maintenance and Repair. Proceed-

ings of the Second International Arch Bridge Conference, A Sinopoli (Ed),

Venice 6.–9. October 1998, Balkema Rotterdam, pp 181–186

Roca P & Molins C (2004) Experiments on Arch Bridges. Arch Bridges IV -

Advances in Assessment, Structural Design and Construction, P Roca &

C Molins (Eds), Barcelona, pp 365–374

Rombach G (2007) Probleme bei der Berechnung von Stahlbetnkonstruktionen

mittels dreidimensionaler Gesamtmodelle. Beton- und Stahlbetonbau 102,

Heft 4, pp 207–214

Rots JG & van Zijl G (2005) Analysis of Concrete and Masonry Structures with

DIANA 9. Weiterbildungsseminar 31.5.-3.6.2005, München

Rouxinol GAF, Providencia P & Lemos JV (2007) Bridgemill bridge bearing ca-

pacity assessment by a discrete element method. Proceedings of the Arch`07 –

5th International Conference on Arch Bridges. PB Lourenco, DV Oliveira &

A Portela (Eds). 12–14 September 2007, Madeira, pp 669–676

Royles R & Hendry AW (1991) Model tests on masonry arches, Proceedings of

the Institution of Civil Engineers, 91 (2), pp 299–321

Rücker W, Hille F & Rohrmann R (2006) Guideline for the Assessment of Exist-

ing Structures. SAMCO Final Report 2006

Ruiz ME, Castelli EA, Prato TA (2008) A new bridge management system for the

National Department of Transportation of Argentina. Bridge Maintenance,

Safety, Management, Health Monitoring and Informatics, H Koh & DM

Frangopol (Eds), Taylor & Francis Group, London, pp 598–605

Schlegel R & Rautenstrauch K (2000) Ein elastoplastisches Berechnungsmodell

zur räumlichen Untersuchung von Mauerwerkstrukturen. Bautechnik 77, Heft

6, pp 426–436

Schlegel R & Will J (2007) Sensitivitätsanalyse und Parameteridentifikation von

bestehenden Mauerwerkstrukturen. Mauerwerk 11, Heft 6, pp 349–355

Schlegel R (2004) Numerische Berechnung von Mauerwerksstrukuren in

homogenen und diskreten Modellierungsstrategien. Dissertation. Bauhaus-

Universität Weimar

Schlegel R (2008) Möglichkeiten der numerischen Simulation von Mauerwerk

heute anhand praktischer Beispiele. Mauerwerkskalender 2009, Ernst und

Sohn

Schlegel R, Will J & Popp J (2003) Neue Wege zur realistischen Standsi-

cherheitsbewertung gemauerter Brückenviadukte und Gewölbe unter

Berücksichtigung der Normung mit ANSYS. 21. CAD-FEM User’s Meeting

2003, 12.–14. November 2003, Potsdam

Schreyer C (1960) Praktische Baustatik – Teil 3., BG Teubner Verlagsgesellschaft

Leipzig

Schueremans L (2001) Probabilistic evaluation of structural unreinforced

masonry, December 2001, Dissertation, Katholieke Universiteit Leuven,

162 3 Computation of Historical Arch Bridges

Heverlee, Belgien,

Schueremans L, Van Rickstall F, Ignoul S, Brosens K, Van Balen K & Van

Gemert D (2003) Continuous Assessment of Historic Structures – A State of

the Art of applied Research and Practice in Belgium. KULeuven, Department

of Civil Engineering

Seim W & Schweizerhof K (1997) Nichtlineare FE-Analyse eben beanspruchter

Mauerwerksscheiben mit einfachen Werkstoffgesetzen. Beton- und

Stahlbetonbau 92, Heft 8, pp 201–207 and Heft 9, pp 239–244

Seim W (1998) Numerische Modellierung des anisotropische Versagens

zweiachsig beanspruchter Mauerwerksscheiben. Aus Forschung und Lehre,

Institut für Tragkonstruktionen, Universität Karlsruhe, Heft 27

SIA 269 (2008) Grundlagen der Erhaltung von Tragwerken.

Slowik V, Fiedler LD & Kapphahn G (2005) Zur kombinierten Anwendung

experimenteller und rechnerischer Methoden bei der Tragsicherheitsbe-

wertung bestehender Massivbauwerke, 3. Symposium Experimentelle

Untersuchungen von Betonkonstruktionen, Technische Universität Dresden,

Institut für Massivbau, 23.Juni 2005, Tagungsband, pp 157–168

Smars (2006) Calipous: http://pierre.smars.dyndns.org/calipous/

Smith CC, Gilbert M & Callaway PA (2004) Geotechnical issues in the analysis of

masonry arch bridges. Arch Bridges IV – Advances in Assessment, Structural

Design and Construction, P Roca & C Molins (Eds), pp 344–352

Stavrouli ME & Stavroulakis GE (2003) Unilateral contact analysis and failure

prediction in stone bridges. Structural Studies, Repairs and Maintenance of

Heritage Architecture VIII, CA Brebbia (Ed), WIT-Press, pp 308–315

Straub H (1992) Die Geschichte der Bauingenieurkunst. Ein Überblick von der

Antike bis in die Neuzeit. Birkhäuser Verlag Basel

Thavalingam A, Bicanic N, Robinson JI & Ponniah DA (2001) Computational

framework for discontinuous modelling of masonry arch bridges. Computers

and Structures, 79, pp 1821–1830

Toker S & Ünay A I (2004) Mathematical modelling and finite element analysis

of masonry arch bridges. G.U. Journal of Science, 17 (2): pp 129–139

Towler KDS (1985) Applications of non-linear finite element codes to masonry

arches, Proceedings of the 2nd International Conference on Civil and Struc-

tural Engineering Computing

Trautz M (1998) Zur Entwicklung von Form und Struktur historischer Gewölbe

aus Sicht der Statik. Dissertation, Institut für Baustatik, Universität Stuttgart,

Stuttgart

UIC-Codex (1995) Empfehlungen für die Bewertung des Tragvermögens

bestehender Gewölbebrücken aus Mauerwerk und Beton. Internationaler

Eisenbahnverband. 1. Ausgabe 1.7.1995

University of Sheffield (2008) http://www.ring.shef.ac.uk/

Vermeltfoort AT (2001) Analysis and experiments of masonry arches. Historical

Constructions, PB Lourenço & P Roca (Eds), Guimarães, pp 489–498

Voigtländer J (1971) Beitrag zur Ermittlung der Schnittkraftumlagerung in

Gewölbebrücken infolge Rissbildung, Dissertation Hochschule für Verke-

hrswesen “Friedrich List” Dresden

References 163

Leipzig

Weber WK (1999) Die gewölbte Eisenbahnbrücke mit einer Öffnung.

Begriffserklärungen, analytische Fassung der Umrisslinien und ein erweitertes

Hybridverfahren zur Berechnung der oberen Schranke ihrer

Grenztragfähigkeit, validiert durch einen Großversuch. Dissertation, Lehrstuhl

für Massivbau der Technischen Universität München

Wedler B (1997) Einfluß der Füllungsschichtung auf das Tragverhalten des

mehrschaligen Mauerwerks in historischen Bauwerken. Diplomarbeit.

Technische Universität Dresden. Lehrstuhl für Massivbau

Weigert A (1996) Untersuchung zum Einfluss der Steinhöhe auf das

Tragverhalten des einschaligen Mauerwerks aus Elbsandstein. Diplomarbeit.

Lehrstuhl für Massivbau. Technische Universität Dresden

Willam KJ & Warnke EP (1974) Constitutive Model for the Triaxial Behaviour of

Concrete, IABSE Report Vol. 19, 1974, Colloquium on „Concrete Structures

Subjected to Triaxial Stresses“. ISMES Bergamot

Witzany J & Jäger W (2005) Die Karlsbrücke in Prag. Mauerwerk 9, Heft 3, pp

108–119

Wolf F (1989) Ein Beitrag zur Beurteilung des Tragverhaltens von massiven

Eisenbahn-Gewölbebrücken. Dissertation, Hochschule für Verkehrswesen.

Dresden

4 Masonry Strength

space.”

4.1 Introduction

In the previous chapter, the numerical modelling of arch bridges was dis-

cussed intensively. During the discussion, it should have become clear that

in most cases, historical arch bridges fail due to the development of a

mechanism chain with hinges, by sliding, or by a combination of this. For

these models, the compression strength is not of utmost importance. How-

ever, in two other cases one has to consider the masonry compression

strength: first looking at the developing hinges where maximum compres-

sion forces are reached, and second considering arch bridges under maxi-

mum equal load. Maximum equal load can be reached by widening the

road lane and therefore increasing the dead load of the bridge.

The estimation of the maximum strength of masonry is not a simple

task. Masonry is a multi component building material. Due to the enor-

mous variety of physical and chemical properties of the components

(Table 4-1), many different numerical models have been developed over

time to describe the strength of masonry. In general, the different numeri-

cal models can be classified into different types of models. The model

classifications are shown in Fig. 4-1.

D. Proske, P. van Gelder, Safety of Historical Stone Arch Bridges, DOI 10.1007/978-3-540-77618-5_4,

© Springer-Verlag Berlin Heidelberg 2009

166 4 Masonry Strength

Table 4-1. Some major factors influencing the masonry strength according to

Wenzel (1997)

Stone and brick Mortar Masonry Structural element

Compression strength Compression Joint thickness Dimensions

Flexural bending strength Joint filling Slenderness

strength Adhesion bond Cavity ratio Support conditions

Stress–strain curve with stone Layer thickness Stiffening

Throatiness Stone order Connection to other

Surface working longitudinal and structural elements

transverse Direction of loading

Eccentricity

et al. (2004)

Stones are static homogenous, natural mineral conglomerates, which make

up a compound caused by some geo dynamical processes (Herrbach 1996).

Stones are classified according to their origin into genetic systems. This

system distinguishes three major groups (Table 4-2). Stones evolve from

mineral melt, magma and lava igneous. Stones originating from some

diagenetic processes of deposited material are called sediment stones.

4.2 Masonry Elements 167

Stones that show new crystallization under high pressure and temperature

are called metamorphic stones.

Major group Sub-group Examples

Igneous Plutonic Granite

Volcanic Basanite

Matrix Gabbro

Sedimentary Clastic sediments Sandstone

Chemical sediment Lime stone

Biogenic sediment Chert

Residual stones

Metamorphic Mica schist

and Hill 1999). Further properties such as colour, structure, or technical

properties have been used to identify the so-called varieties. Peschel

(1984) clearly shows the differences of these technical properties. Techni-

cal properties and possible application of natural stones are given in Peschel

(1984), Dienemann and Burre (1929), Gäbert et al. (1915), and Schubert

(2004) for German natural stones. A summary is listed in Table 4-3.

Table 4-3. Technical properties of some natural stones according to Stein (1993)

Weights Compression Flexural bend- Thermal expan-

in kN/m3 strength in ing strength in sion coefficient

MPa MPa in mm/mK

Granite, Syenite 28 160–240 10–20 0.80

Diorite, Gabbro 30 170–300 10–22 0.88

siliceous porphyry 28 180–300 15–20

Basalt lava 24 80–150 8–12

Diabase 29 180–250 15–25 0.75

Quartzite, Greywacke 27 150–300 12–20

Siliceous sandstone 27 120–200 3–15

Dense lime and dolomite 28 80–180 6–15 0.75

stones

Other lime stones 28 20–90 5–8

Travertine 26 20–60 4–10 0.68

Volcanic Tuff 20 2–6

Gneiss 30 160–280 10–15

factors that have to be considered and stated. Such factors are, for exam-

ple, the size of the test specimen, the humidity of the stone material, or the

168 4 Masonry Strength

layering of the stone material. Figure 4-2 illustrates the effect of the

specimen size on the flexural bending strength of granite stone in terms of

specimen height. The left side of the diagram shows data that have been

taken from many different references, whereas on the right side predomi-

nantly its own data have been used. Furthermore, on the far right, data

from uniaxial tensile tests have been included. A complete list of data ref-

erences is given in Curbach and Proske (2003).

Fig. 4-2. Flexural bending strength of granite based on the test specimen height

(Curbach et al. 2004)

the stone properties, but natural stone material also shows, in most cases,

strong deviation from the mean value. This is exemplarily shown in Figs.

4-3 and 4-4. Here, compression and splitting tensile test results are shown

as frequencies for 500 samples of Saxon sandstone (from Lohmen). Stud-

ies by Curbach and Proske (1998) indicate a possible application of the

normal distribution (Gauss distribution) or the beta distribution for the

compression strength of natural stones. Furthermore, Fig. 4-5 shows a corre-

lation analysis for the data. From this figure, it becomes clear that a correla-

tion between the aforementioned strength values is almost negligible.

Further, such investigations are shown in Proske (2003). For German natural

stones, Peschel (1984) gives a nearly complete list of technical properties

including not only mean values but also measurements for variance and

deviation.

4.2 Masonry Elements 169

Sometimes, not only the maximum strength values of the natural stones

are required but also the complete stress-strain relationship is needed.

Although this topic will not be discussed here in detail, an example of a

stress-strain relationship is shown in Fig. 4-6 for Silesia sandstone. More

detailed information can be found in Alfes (1992) for sandstone.

Fig. 4-3. Histogram for the compression strength of Saxon sandstone according to

Curbach and Proske (1998)

170 4 Masonry Strength

Fig. 4-4. Histogram for the splitting tensile strength of Saxon sandstone according

to Curbach and Proske (1998)

splitting tensile strength according to Curbach and Proske (1998)

4.2 Masonry Elements 171

The same types of stones are frequently used within a region. This effect is

often visible if one compares geological maps, such as the one for the Free

State Saxony (1992), with the geographical distribution of certain natural

building materials.

This is especially true for sediment stones, which were heavily used in

former centuries. This type of stone is easy to work with. Therefore, for

example in Saxony, many buildings received a covering of sandstone.

Igneous stones were also used in early times for structures. Granite, for

example, was already used as a building material in Germany by the Ro-

mans in the 2nd century (Müller 1977). In contrast to the sediment stones,

working with igneous stones was much more difficult. Therefore, different

working levels exist for this type of stone.

Quarry stones and cobblestones are usually rough and irregular in ge-

ometry. There are great deviations in the size of these stones: the size de-

pends on their strength and workability. The only joints observed by geo-

logical processes acting on the stones themselves are parallel horizontal

joints, and such joints can be bedding.

Cut stones show abrasive worked joints where the stones are mainly

produced by cleaving. Horizontal and vertical joints are already in a rec-

tangular angle. The chosen geometry still depends very much on the stone

strength. Freestones or cut stones feature sight surfaces that are manufac-

tured based on geometrical and artistic requirements.

The sight surfaces can furthermore be distinguished as shown in Table

4-4 and Fig. 4-7.

172 4 Masonry Strength

Table 4-4. Working types of stone sight surfaces according to Warnecke (1995)

Name Pointed Flattened Pilled Charring Wide charring

1st action 2nd action

Muster

Tool

Time Until middle Until End 12th until Middle 15th From middle

of the 11th beginning of End 13th to end 17th 17th century

century the 12th century century

century

Fig. 4-7. Further stone sight surface working types, such as the Belgium and the

Dutch type, according to Van der Vlist et al. (1998)

4.2.2 Mortar

Historically, masonry arch bridges can be found with and without mortar.

Masonry can tolerate some loads even if it is fabricated with sand instead

of mortar or without any joint material (this will be mentioned again later).

However, historical bridges constructed without mortar usually had iron

clamps to provide connection to the next stone (Straub 1992). An example

of a historical bridge without masonry is the Pont du Gard in France

(Garbrecht 1995). This bridge clearly shows that such structures can

survive long lifetimes. Armaly et al. (2004) assume that the iron clamps

can increase the load-bearing capacity up to 30%.

Besides those bridges without mortar, most historical arch bridges made

of masonry have used, and still consist of, mortar. The properties of his-

torical mortar are discussed in many publications such as in Baronio and

Binda (1991), Huesmann and Knöfel (1991), Knöfel and Schubert (1991),

Knöfel and Middendorf (1991), Wisser and Knöfel (1988), Warnecke

(1995), Franken and Müller (2001), Freyburg (1994), Franken (1995),

Gucci and Barsotti (1995), van Hees et al. (2004), Domède et al. (2008),

and Bökea et al. (2006). Figure 4-8 gives an overview about the many dif-

ferent factors influencing the final strength of the mortar.

4.2 Masonry Elements 173

modern mortar. Tables 4-5 and 4-6 show some values of the compression

strength of historical mortar. Figure 4-9 shows the stress-strain relationships

of different mortar types, and Fig. 4-10 gives a comparison of historical

and modern mortars.

Fig. 4-8. Influences on the strength of mortar inside masonry according to Huster

(2000)

Stefanidou (2003)

Structure Period Compression strength in MPa

Roman Forum 2nd century 2.5–4.0

Galerius Palace 3rd century 3.0–4.5

Acheropiitos 5th century 2.3–3.0

Hagia Sophia 7th century 2.0–6.0

Hagios Panteleimonas 14th century 1.0–1.4

Hagia Aikaterini 13th century 1.6–2.0

Bezesteni 16th century 2.5–3.5

Old house Mouson 19th century 1.5–2.0

Table 4-6. Compression strength of historical mortars according to the COST 345

(2006)

Mortar class Mixture ratio cement : lime : sand Compression strength in MPa

(volume)

I 1:0–0.25:3 11–16

II 1:0–5:4.5 4.5–6.5

III 1:1:5–6 2.5–3.6

IV 1:2:8–9 1.0–1.5

V 1:3:10–12 0.5–1.0

VI 0:1:2–3 (hydraulic lime) 0.5–1.0

VII 0:1:2–3 (pure lime) 0.5–1.0

174 4 Masonry Strength

Fig. 4-9. Comparison of the stress-strain relationship for historical and modern

mortar according to Warnecke (1995)

Frenzel (2004)

cause failure of historical structures. An example of the failure of a struc-

ture where mortar could potentially have been the cause was the city tower

in Pavia. However, mortar did not cause the failure of this medieval tower,

resulting in four fatalities, as studies showed later. Other examples of the

failure of historical structures in relation to the creep of masonry, with a

significant contribution by mortar, are given in Verstrynge et al. (2008).

Furthermore, the strength of historical mortar shows great deviations as

illustrated in Figs. 4-11 and 4-12. Figure 4-11 shows the distribution of the

mortar strength for the city tower in Pavia. The data from Fig. 4-12

originate from a bridge built in 1875. In the face of statistical analysis, it

should be mentioned here that very often the mortar data from historical

structures are heavily censored. This means that very often mortar cores or

specimens do not survive the exploitation process and only the strongest

4.2 Masonry Elements 175

can be finally tested. This should be considered and kept in mind while in-

interpreting mortar data from such structures. The deviation of the mortar

strength is related not only to the exploitation process but also to the loss

of binder and to the grading curves as shown in Fig. 4-13.

For the testing of mortar, many different codes and recommendations

exist, such as DIN 18555 1-9 (1982).

Baronio and Binda (1991)

176 4 Masonry Strength

measurements of one bridge by Proske (2003)

Fig. 4-13. Comparison of the grading curves of some historical (left) and some

Roman mortar (right) according to Wisser and Knöfel (1988)

Based on the known properties of the single elements of masonry, the es-

timation of the properties of the masonry should be theoretically possible.

Although this trial has been carried out frequently, an entire theory about

the estimation of the strength properties based on the mechanical properties

of the elements is still missing. Most models are based on simple empirical

4.3 Maximum Centric Masonry Compression Strength 177

the investigated masonry type. A chronological list of models partially

taken from Purtak (2001), Schulenberg (1982), and Mann (1983) for the

maximum masonry compression strength is given by the following:

Krüger (1916), Graf (1926), Drögsler (1933), Voellmy (1937), Drögsler

(1938), Hansson (1939), Hermann (1942), Kreüger (1943), Nylander

(1944), Svenson (1944), Haller (1947), Ekblad (1949), Oniszczyk (1951),

Bröcker (1961), Hilsdorf (1965/69), Monk (1967), Francis et al. (1970),

Khoo and Hendry (1972), Brenner (1973), Schnackers (1973), Kirtschig

(1975), Probst (1981), Schulenberg (1982), Rustmeier (1982), Mann

(1982/83), Atkinson et al. (1985), Ohler (1986), Berndt (1992/96), Sabha

and Pöschel (1993), Babylon (1994), and Ebner (1996).

However, the simple naming of these models is not sufficient because

most models are only based on the analysis of some tests, whereas other

models include some theoretical considerations. The models based on

limited tests are strongly related to special conditions or properties of the

tests, such as stone type. Therefore, in the following, some of the models

will be discussed in more detail.

The new German code of practice DIN 1053-100 (2004) is the follower of

the DIN 1053-1 (1996). The new code gives some rough measures for the

natural stone masonry compression strength based on the stone strength

and the mortar type. The application is simple since only two tables have

to be used (Tables 4-7 and 4-8). First, the masonry has to be classified, and

second, according to the classification, the stone strength, and mortar class,

the masonry strength can be estimated.

Quality General classification Joint height to Angle of Transfer factor η

category stone length joint in tan

α

N1 Quarry stone masonry ≤ 0.25 ≤ 0.30 ≥ 0.50

N2 Hammered coursed rubble ≤ 0.20 ≤ 0.15 ≥ 0.65

masonry

N3 Coursed rubble masonry ≤ 0.13 ≤ 0.10 ≥ 0.75

N4 Ashlar masonry ≤ 0.07 ≤ 0.05 ≥ 0.85

178 4 Masonry Strength

Table 4-8. Characteristic mortar compression strength based on stone strength and

mortar class according to DIN 1053-100 (2004)

Quality category Stone compression Mortar compression strength fk in MPa

strength fbk subject to the mortar group

I II IIa III

N1 ≥ 20 Mpa 0.6 1.5 2.4 3.6

≥ 50 Mpa 0.9 1.8 2.7 4.2

N2 ≥ 20 MPa 1.2 2.7 4.2 5.4

≥ 50 MPa 1.8 3.3 4.8 6.0

N3 ≥ 20 MPa 1.5 4.5 6.0 7.5

≥ 50 MPa 2.1 6.0 7.5 10.5

≥ 100 MPa 3.0 7.5 9.0 12.0

N4 ≥ 5 MPa 1.2 2.0 2.5 3.0

≥ 10 MPa 1.8 3.0 3.6 4.5

≥ 20 MPa 3.6 6.0 7.5 9.0

≥ 50 MPa 6.0 10.5 12.0 15.0

≥ 100 MPa 9.0 13.5 16.5 21.0

The former German code of practice DIN 1053 (1996) also gave a table for

the estimation for the natural stone masonry compression strength (Table

4-9). It is likely that the table was based on the model from Mann, because

the stone compression strength has only a minor influence on the masonry

compression strength, and the stone tensile strength has not been considered.

Table 4-9. Characteristic mortar compression strength based on stone strength and

mortar class according to DIN 1053 (1996)

Quality Stone compression Mortar compression strength σ0 in MPa subject

category strength fbk to the mortar group

I II IIa III

N1 ≥ 20 MPa 0.2 0.5 0.8 1.2

≥ 50 MPa 0.3 0.6 0.9 1.4

N2 ≥ 20 MPa 0.4 0.9 1.4 1.8

≥ 50 MPa 0.6 1.1 1.6 2.0

N3 ≥ 20 MPa 0.5 1.5 2.0 2.5

≥ 50 MPa 0.7 2.0 2.5 3.5

≥ 100 MPa 1.0 2.5 3.0 4.0

N4 ≥ 20 MPa 1.2 2.0 2.5 3.0

≥ 50 MPa 2.0 3.5 4.0 5.0

≥ 100 MPa 3.0 4.5 5.5 7.0

4.3 Maximum Centric Masonry Compression Strength 179

The tables mentioned in the German codes are a very simple way to assess

the compression strength of natural stone masonry. Pure empirical models

of the masonry compression strength fmas,c have also found wide application

due to their simple application. Furthermore, they are easy to develop by

simple regression analysis. A common type are exponential equations

using the stone compression strength fst,c and the mortar compression

strength fmo,c such as developed by Schubert and Krämer:

fmas,c = a ⋅ fstb,c ⋅ fmo

c

,c .

(4-1)

This type is, for example, used in the Eurocode 6. The 5% quantile of

the masonry compression strength fmas,c,k is then computed using average

compression strength values of the stone fst,c,m and the mortar fmo,c,m:

fmas,c,k = 0.40 ⋅ fst0.75

,c , m ⋅ f mo ,c , m .

0.25 (4-2)

fmas,c,m = 0.83 ⋅ fst0.66

,c , m ⋅ f mo ,c , m .

0.18 (4-3)

However, Mann’s parameters may not fit very well for natural stone

masonry. Therefore, the Ril 805 (1999) suggests the following exponents:

fmas,c ,m = 0.80 ⋅ fst0.70

,c , m ⋅ f mo ,c , m .

0.20 (4-4)

oped a model for the estimation of the compression strength of masonry

based on the multi axial stress conditions in the stone and the mortar. The

model therefore includes some theoretical considerations. In general, the

model assumes that the low Young’s modulus of the mortar restrains the

deformation of the stone posited in the masonry. These restraints cause

transverse tensile forces inside the stone and transverse compression forces

inside the mortar (Fig. 4-14). First, Hilsdorf simply assumed that there was

a perfect bond between the mortar and the stone, but later dismissed this

assumption. The model was primarily developed for brick masonry, yet

was later adapted to natural stone masonry by a so-called asymmetry factor.

The major advantage of this model is the theoretical consideration, but a

180 4 Masonry Strength

1997, and Warnecke et al. 1995).

The masonry compression strength is given as

fst ,c (4-5)

fst ,sp + a ⋅ fst ,c

strength, fst,sp as stone splitting tensile strength, fmo,c as mortar compression

strength, u as asymmetry factor, and a as

t (4-6)

a= h

4.1

with t as joint height and h as stone height.

Fig. 4-14. Stress stages inside the masonry according to Hilsdorf (Warnecke et al.

1995)

stones (bricks) differs significantly from that of natural stones. The

asymmetry and roughness of the stones and the joints yield to a

qualitatively different load-bearing mechanism. Furthermore, natural stones

mainly show a higher tensile strength compared to bricks. Therefore,

4.3 Maximum Centric Masonry Compression Strength 181

Mann assumes that the failure of natural stone masonry will be dominated

by the failure of the mortar inside the masonry. This assumption, however,

is in complete contradiction of test results with masonry built with sand in-

instead of mortar. Usually, the uniaxial compression strength of the sand is

virtually negligible. Then, the masonry is still able to take considerable

loads whereas the formula from Mann would give the masonry

compression strength of zero. The masonry failed by tearing of the stones

(Warnecke et al. 1995).

f mas ,c = f mo ,c ⋅ f ⋅ ü (4-7)

8 1 (4-8)

f = ⋅ 2

9 ⎡ 2 t⎤

1 − ⎢1 − ⋅ ⎥ ⋅ cos 4 α

⎣ 3 b⎦

AS (4-9)

ü=

AMW

In this, b is the width of the stone, ü is the effective cross section and α

is the angle of the joints. Mann’s formula gives good results for rubble

masonry with weak mortar. It can be related to the works by Rustmeier

(1982).

Berndt (Berndt 1996, Berndt and Schöne 1991, and Wenzel 1997) has

developed a concept for coursed rubble masonry from the so-called Elbe-

sandstone. Berndt assumes a splitting tensile failure of the stone. As an

extension to the work from Hilsdorf, Berndt not only considers tensile

forces inside the stone due to constrained deformation of the mortar caused

by the stone, but also tensile forces caused by force direction changes due

to the unequal cross-sectional areas of the mortar and of the stone. The

estimation of the maximum masonry compression strength is given as

f st ,c (4-10)

f ma ,c =

⎡t v b d ' ⎤ f st ,c

⎢⎣ h ⋅ 1 − v + k ⋅ h ⋅ b ⎥⎦ ⋅ f + 0.7

st , sp

with

k = 0.3...0.5 (4-11)

182 4 Masonry Strength

t (4-12)

d'≈t+

⎛ ρ⎞

tan ⎜ 45 + ⎟

⎝ 2⎠

and

⎧ h ⎫ (4-13)

h ' = min ⎨ ⎬.

⎩10 cm ⎭

Besides the shear formula, Berndt and Schöne (1991) have furthermore

introduced a safety concept for the application of this formula. This is very

useful, since for many formulas it is unclear what characteristic value of

the masonry compression strength is computed. Table 4-10 shows the

single elements of this concept.

Table 4-10. Safety elements in the safety concept for the evaluation of the

compression masonry strength using the model by Berndt and Schöne (1991)

Factor Description

m1 Considers the change from mean masonry compression strength to

characteristic masonry strength, usually the 5% fractile value

m2 Considers the slenderness oft the test specimen

m3 Considers the impossibility of load flow changes in piers

m4 Considers the change from mean stone compression strength to charac-

teristic stone compression strength, usually the 5% fractile value:

s

m4,1 = 1 − 1.645 ⋅

fst ,c

Considers the change from mean stone splitting tensile strength to cha-

racteristic stone splitting tensile strength, usually the 5% fractile value:

s

m4,1 = 1 − 1.645 ⋅

fst , sp

m5 Considers the joint thickness influence on the load-bearing behaviour:

m5 = 0.85

m6 Considers the Sprödbruch behaviour of the masonry: m6 = 0.85

m7 Considers the influence of the stone layering on the masonry

compression strength: m7 = 0.90

m8 Considers the long-term loading strength of the masonry: m8 = 0.90

4.3 Maximum Centric Masonry Compression Strength 183

m1 = 0.426 ⋅ 0.604 ⋅ 0.85 ⋅ 0.85 ⋅ 0.90 ⋅ 0.90 = 0.151 .

The characteristic masonry compression strength then reaches

fma,c ,k = fma,c,m ⋅ ∏ m = 30.9 ⋅ 0.151 = 4.66 MPa (4-15)

The models of Berndt and Sabha have both been developed in Dresden,

Germany, and are both strongly connected to the Elb sandstone found in

this region. Furthermore, both authors consider mechanisms that cause a

splitting tensile failure of the stones when masonry fails under maximum

compression forces. However, as an extension to Berndt, Sabha (Sabha

and Schöne 1994, Sabha and Weigert 1996, and Wenzel 1997) considers

the locations of regions with maximum splitting tensile forces for the two

mechanisms causing such forces. Whereas the maximum tensile force due

to force change direction is approximately in the middle of the stone

height, the maximum tensile force due to strain restraints of the mortar is

reached in the stone heights close to the mortar. Therefore, Sabha does not

add both the tensile forces that should reach higher masonry compression

forces in comparison to Berndt:

2 ⋅ k ⋅ f st ,c + f st ,sp (4-16)

f ma ,c =

f

k + st ,sp

f st ,c

with

t⎛ f st , sp ⎞ (4-17)

k = 1.6 ⎜⎜ 1.45 + 1⎟⎟ .

b⎝ f st ,c ⎠

Boye (1998) gives an extension of the Sabha model for flat stones.

historical railway bridges uses a model that is based on works by Ohler

(1986). The formula of Ohler (1986) is given as

184 4 Masonry Strength

fma ,c = 0.5 ⋅ fmo ,c ,m +

b ⋅ hF ⋅ 0.5 ⋅ fst ,c ,m

1+

2 ⋅ hS ⋅ 0.05 ⋅ fst ,c ,m

with hF as mortar joint thickness and hs as stone height. The splitting

tensile strength of the stones inside the formula is considered as 5%

fractile value.

developed a simple model that only considers the density of the stones γ

and the mortar quality in terms of mortar groups (MG). The model is

dominated by the stone failure according to Huster (2000).

0.007 ⋅ (18.7 ⋅ γ − 355.2 MPa) for MG I (4-19)

fma = 0.017 ⋅ (18.7 ⋅ γ − 355.2 MPa) for MG II .

0.024 ⋅ (18.7 ⋅ γ − 355.2 MPa) for MG III

2001 and Simon 2002). The masonry compression strength is computed as

1 (4-20)

fma ,c = fst ,c ⋅

fst ,c ⎛ Est ⎞

⋅⎜ ⋅ μmo − μst ⎟

1 + st ,sp ⎝ mo ⎠ .

f E

hS E st

⋅ ⋅ (1 − μmo )

t E mo

The model of Khoo and Hendry (1972) uses a cubic equation for the

estimation of the failure curves of stones and masonry. The masonry

compression strength is then given by

4.3 Maximum Centric Masonry Compression Strength 185

hs (4-21)

(0.997 ⋅ fst ,sp + 0.162 ⋅ ⋅ fmo ,c ) +

t

fst ,sp h

(0.203 ⋅ + 0.113 ⋅ s ) ⋅ fmas ,c +

fst ,c t

fst ,sp hs

(1.278 ⋅ − 0.053 ⋅ ) ⋅ fmas

2

,c +

f 2

st , c t ⋅ fmo ,c

fst ,sp hs

(0.249 ⋅ − 0.002 ⋅ ) ⋅ fmas

3

,c = 0 .

fst3,c t ⋅ fmo

2

,c

The works by Khoo and Hendry were extended by Probst (Simon 2002).

hs (4-22)

1 ⋅ fst ,sp + t ⋅ fmo ,sp

fmas ,c = ⋅ 2 .

μmas hs + t

⎡ t ⎛ σ y ⎞ ⎤ ⎛ n ⋅ bs ⎞

0.4 (4-23)

= ⎢1 − 1.2 ⋅ ⋅ (1 − 2 ⋅ tan ϕ ) ⋅ ⎜ ⎟ ⎥ ⋅ ⎜1 −

l ⎟⎠

fmas,c .

⎢⎣ d ⎝ c ⎠ ⎥⎦ ⎝

are known as mentioned in the beginning of this chapter. Such models, not

discussed here, are (for example) models by Atkinson et al. (1985),

Rustmeier (1982), or Pöschel. The comparison of all models based on

different items such as model deviation, robustness, convergence, possible

measurement of the input data, and minimum of required input data

would exceed the capacity of this book. For the interested reader, the

186 4 Masonry Strength

Budelmann (1995) can be recommended.

estimation of masonry structural elements. Usually, structural elements not

only are exposed to axial forces but also have to bear moments and shear

forces. For such investigations, usually the stress-strain relationship for

masonry under axial forces is required.

Several models of such relationships can be found in literature. An

overview of current models has been given by Glock (2004), Lissai (1986),

Becker and Bernard (1991), and Walthelm (1990). Glock lists the

following models:

• Angervo (mineralic no-tensile materials)

• Becker and Bernard (masonry)

• Lewicki (concrete)

• Sargin (concrete)

• Jäger (masonry)

• DIN 1045 (concrete)

• Eurocode 6 (masonry), see Fig. 4-15.

A second look at the models reveals that only a minority is related to

masonry and most stress-strain relationships originate from concrete.

tions of the ultimate load-bearing behaviour of masonry structural ele-

ments offers an increase in the numerical load by up to 25%, according to

Becker and Bernard (1991). Figure 4-16 shows the development of different

4.5 Moment-Axial Force Diagrams 187

force with eccentricity.

Fig. 4-16. Different stress distribution in a cross section according to Mann (1991)

application of stress-strain relationships, moment-axial force diagrams can

also be used. The advantage of this diagram is the consideration of the

nonlinear behaviour of the stress-strain relationship of masonry in a simple

way, while the disadvantage is numerous computations to prepare such

diagrams. However, if such diagrams are available, they can be easily used

by practitioners. Such diagrams have been developed and published by

Purtak (2001) – Fig. 4-17 for masonry walls and for arch bridges (Purtak et

al. 2007). Furthermore, Lissai (1986) and Pauser (2005) have also prepared

such moment-axial force diagrams.

Fig. 4-17. Example of a moment-axial force diagram from Purtak et al. (2007)

188 4 Masonry Strength

4.6.1 Introduction

consists in most cases of additional leafs due to the significant thickness of

the masonry structural elements. This is also true for the piers of arch

bridges or other elements of the arch bridges. Such a multi-leaf structure

can often be proven by horizontal drillings into the piers.

Again, there exist many different models for such multi-leaf masonry.

An introduction this field is given in Warnecke et al. (1995). In this

chapter, only the models by Warnecke and Egermann are introduced.

maximum forces for multi-leaf masonry elements. He assumes that a

correct estimation of the strength of masonry elements alone from drillings

is not possible. Furthermore, a cohesive inner layer is considered. The

following formulas show further assumptions, such as

vMo + vSt + v Hohlraum = 1 (4-24)

= + .

Ei Emo ⋅ vmo ESt

The strength of the inner masonry layer can be computed as

vmo (4-26)

f mas ,c ,i = f mo ,c ⋅ .

1 − vSt

The strength of the outer masonry layers can be done equal to single-

leaf masonry.

tions:

• External leaf with brick masonry and stretcher bond, the slenderness is

less than 13.3

4.6 Additional-leaf Masonry 189

• There exists a plain surface between the external and the internal leaf

• Bernoullis hypothesis is valid (even strain distribution of cross section)

• Symmetrical support conditions at base and crown (usually fixed)

• Rigid foundation for the entire cross section

The basis value can be evaluated according to a singe-leaf masonry.

However, this value has to be adapted according to

f DA = α λ ⋅ αϕ ⋅ f mas ,c (4-27)

with

fDA Masonry compression strength of the external leaf

fmas,c Masonry compression strength of the external leaf computed as

single-leaf masonry

αϕ Consideration of the direction of pre-stressing

αϕ = 1 pre-stressing direction parallel to the loading direction

αϕ = 2 pre-stressing direction rectangular to the loading direction

AA Cross section of the external leaf

I Moment of inertia for the non-cracked cross section

0.7 Decrease factor for cracking

sk Effective length

1 (4-28)

α λ = 1 for N cr ≥ N W ,0

2

N cr 1 (4-29)

αλ = 2 ⋅ for N cr < NW ,0

NW ,0 2

NW ,0 = AA ⋅ σ D , MW (4-30)

E⋅I (4-31)

N cr = 0.7 ⋅ π 2 ⋅

sk2

E ≈ 1,000 ⋅ σ D , MW (4-32)

Finally the masonry compression strength for the overall cross section

can be computed as

190 4 Masonry Strength

AA1 A A (4-33)

f mas = 0.75 ⋅ f mas ,c ,1 + 0.75 ⋅ f mas ,c,2 A 2 + 1.3 ⋅ f mas ,c,3 I .

A A A

As the last equation clearly shows for the external leafs, the

compression strength is decreased compared to a single-leaf masonry and

for the internal leaf, it is increased due to multi axial compression state.

However, practice has shown that the computation under common

conditions does not yield to a significant change in the compression

strength compared to a single-leaf cross-section assumption.

bridges may not only fail due to the development of hinges and chains, but

sliding can also occur in the arch itself. To evaluate the permitted shear

stresses inside the masonry, the failure surfaces discussed in Chapter 3 can

be used. However, it is often desired to apply a more simple proof

comparable to the computation of the maximum compression strength of

the masonry.

Although Mann and Müller (Baier 1999) have developed an excellent

theory for the shear failure of natural stone masonry, the approach here by

Berndt (1996) will be recommended since this approach permits a

continuous technique in combination with the model for the computation

of the maximum compression force.

Identically to Mann and Müller (Baier 1999), Berndt (1996) has

introduced three regions of failure. These three regions can also be

compared to the shear failure of concrete beams.

The first region is simply the Coulombs friction:

τ = f HS + μ ⋅ σ x (4-34)

comparable situation in reinforced concrete is the tensile tie failure under

shear force with an insufficient amount of reinforcement. The shear force

forms a plateau and can be computed with

1 + kσ (4-35)

fst ,c 2

max τ = ⋅ .

1.4 ⎛ f ⎞ ⎛ f ⎞

⎜⎜

st , c

+ 0.7 ⋅ kσ ⎟ ⋅ ⎜ st ,c ⋅ kσ + 0.7 ⎟

f ⎟ ⎜f ⎟

⎝ st , sp ⎠ ⎝ st ,sp ⎠

4.8 Proof Equations 191

Finally, the third region describes the failure of the stones by compres-

sion. This can be compared to the compression strut failure in concrete

beams under high shear forces and high shear force reinforcement:

⎛ f st ,c ⎞ (4-36)

⎜ + 0.7 ⋅ kσ ⎟⎟ ⋅ f st ,c

1 ⎛ 1.4 ⋅ τ ⎞ ⎜⎝ f st , sp

2

f mas ,c ≈

f st ,c

− ⋅⎜ ⎠

⎟ ⋅ .

f st ,c 2 ⎜⎝ f st ,c ⎟⎠ 1 + kσ

⋅ kσ + 0.7

f st , sp 2

If the three equations are used to construct a failure curve, the following

figure can be drawn (Fig. 4-18).

Fig. 4-18. Failure curve for sandstone masonry under shear and axial forces

Very often, either the minimum or the maximum shear forces are under

discussion. According to the German code DIN 1053-1, only a maximum

shear stress of 0.3 MPa can be applied. However, other works have shown

that even the 5% fractile values of the maximum shear strength can reach

values up to 2 or 3 MPa (Baier 1999).

The computed stresses can be used for static proofs in the limit state of the

ultimate load, and in the limit state of serviceability. However, the proof

concepts differ significantly according to the different generations of codes

of practice (Table 4-11). This is mainly based on different safety concepts

as later discussed in Chapter 7.

192 4 Masonry Strength

Table 4-11. Different proof concepts for structural elements under axial forces

Code of practice Loading ≤ resistance

EC 6 N d ≤ Rd

DIN 1053-2 γ ⋅σ R ≤ βR

DIN 1053-1 (Feb. 1990) σ ≤ zul σ D

DIN 1053-100 N d ≤ Rd

bridges. Besides that, the German railway does not recommend the appli-

cation of nonlinear computations of the arch under serviceability. Interna-

tional codes, such as a former version of the Eurocode, with some special

remarks concerning proofs of historical arch bridges even for the limit

state of serviceability, can be found:

• The stress in the extreme fibre should not exceed 65% of the maximum

compression strength

• The computed deformations of the arch under the traffic load at the

vertex (crown) should not exceed 1/1000 of the arch span

The British BABTIE draft (Jackson 2004) recommends for the

serviceability proof:

• Crack depth lower than 0.25 × h

• Stress lower than 0.4 × fk

• No tensile forces under torsion and quasi-permanent loads

If proofs cannot be fulfilled for historical structures, in many cases it

does not mean that the structure shows insufficient safety. One has to

consider that the safety concepts as the basis of codes are mainly

concerned with modern structures. Therefore, as already mentioned in

Chapter 1, the safety concept may be altered, for existing structures. This

statement does not mean that historical structures or bridges can be less

safe, however the applied tools can differ.

References

Sandsteinen. Jahresberichte Steinzerfall – Steinkonservierung 1992, Verlag

Ernst & Sohn, Berlin

Armaly M, Blasi C & Hannah L (2004) Stari Most: rebuilding more than a historic

bridge in Mostar. Museum International, Blackwell Publishing, No. 224, Vol.

56, No. 4, pp 6–17

References 193

Atkinson RH, Noland JL & Abrams DP (1985) A Deformation Failure Theory for

Stack-Bond Brick Masonry Prisms in Compression. Proceedings of the 7th

International Brick/Block Masonry Conference, Melbourne

Baier G (1999) Statistische Auswertung von Schubversuchen an Sandsteinmauerwerk.

Diplomarbeit, Technische Universität Dresden, Lehrstuhl für Massivbau

Baronio G & Binda L (1991) Experimental approach to a procedure for the

investigation of historic mortars. Proceedings of the 9th International

Brick/Block Masonry Conference, 13th–16th October 1991, Vol. 1, Berlin,

Germany, pp 1397–1405

Becker G & Bernard R (1991) Nichtlineare Spannungs-Dehnungs-Verläufe bei

Mauerwerk. Bautechnik 68, Heft 5, pp 5–15

Berndt E & Schöne I (1991) Tragverhalten von Natursteinmauerwerk aus

Elbesandstein. Sonderforschungsbereich 315, Universität Karlsruhe, Jahrbuch

1990

Berndt, E (1996) Zur Druck- und Schubfestigkeit von Mauerwerk – experimentell

nachgewiesen an Strukturen aus Elbsandstein. Bautechnik 73, Heft 4,

pp 222–234

Bökea H, Akkurtb S, İpekoğlua B & Uğurlua E (2006) Characteristics of brick

used as aggregate in historic brick-lime mortars and plasters. Cement and

Concrete Research, 36 (6), June 2006, pp 1115–1122

Börner K & Hill D (1999) Lexikon der Natursteine, CD-ROM, Abraxas Verlag

Boye A (1998) Einfluss der Steinbreite auf die Tragfähigkeit des einschaligen

Mauerwerks. Diplomarbeit, Technische Universität Dresden, Lehrstuhl für

Massivbau

COST 345 (2006) European Commission Directorate General Transport and

Energy: COST 345 – Procedures Required for Assessing Highway Structures:

Working Group 6 Report on remedial measures for highway structures.

http://cost345.zag.si/Reports/COST_345_WG6.pdf

Curbach M & Proske D (1998) Abschätzung des Verteilungstyps der

Mauerwerksdruckfestigkeit bei Sandsteinmauerwerk. Jahresmitteilungen

1998, Schriftenreihe des Institutes für Tragwerke und Baustoffe, Heft 7, TU

Dresden

Curbach M & Proske D (2003) Gutachten zur Ermittlung von Sicherheit-

selementen für die Biegetragfähigkeit von Granit-Steindeckern, Technische

Universität Dresden

Curbach M, Günther L & Proske D (2004) Sicherheitskonzept für Brücken aus

Granitsteindeckern. Mauerwerksbau, Heft 4, pp 138–141

Dienemann W & Burre O (1929) Die nutzbaren Gesteine Deutschlands und ihre

Lagerstätten mit Ausnahme der Kohlen, Erze und Salze. II. Band: Feste

Gesteine, Verlag von Ferdinand Enke, Stuttgart

DIN 1053-1 (1996) Mauerwerk – Teil 1: Berechnung und Ausführung. 11/1996

DIN 1053-100 (2004) Mauerwerk: Berechnung auf der Grundlage des

semiprobabilistischen Sicherheitskonzeptes. NABau im DIN/Beuth Verlag:

Berlin, 7/2004

DIN 18555 1-9 (1982) Prüfung von Mörteln mit mineralischen Bindemitteln.

12/1982

194 4 Masonry Strength

masonry. Materials and Structures. DOI 10.1617/s11527-008-9372-z

Ebner B (1996) Das Tragverhalten von mehrschaligem Bruchsteinmauerwerk im

regelmäßigen Schichtenverband. Berichte aus dem konstruktiven In-

genieurbau, Heft 24, TU Berlin

Eggermann R (1995) Tragverhalten mehrschaliger Mauerwerkskonstruktionen.

Dissertation, Aus Forschung und Lehre 29, Institut für Tragkonstruktionen,

Universität Karlsruhe

Francis AJ, Horman CB & Jerrems LE (1970) The Effect of Joint Thickness Other

Factors on the Compressive Strength of Brickwork, 2. International Brick

Masonry Conference, Stoke-on-Trent, England

Franken S & Müller HS (2001) Historische Mörtel und Reparaturmörtel.

Untersuchen, Bewerten, Einsetzen. Empfehlungen für die Praxis, Fritz

Wenzel und Joachim Kleinmanns (Hrsg) “Erhalten historisch bedeutsamer

Bauwerke”. Karlsruhe

Franken S (1995) Die Frauenkirche zu Dresden. Zur Entwicklung geeigneter

Mörtel für den Wiederaufbau eines historischen Bauwerkes. In: Jahrbuch

(1993) des Sonderforschungsbereiches 315 “Erhalten historisch bedeutsamer

Bauwerke”, 1995, pp 93–110

Frenzel M (2004) Untersuchung zur Quertragfähigkeit von Bogenbrücken aus

Natursteinmauerwerk. Diplomarbeit Technische Universität Dresden

Freyburg E (1994) Bausteine und Mörtel historischer Eisenbahnbrücken zwischen

Weißenfels und Weimar. Wissenschaftliche Zeitschrift der Hochschule für

Architektur und Bauwesen Weimar, 40, 5/6/7, pp 75–77

Gäbert C, Steuer A & Weiss K (1915) Die Nutzbaren Gesteinsvorkommen

Deutschlands. Verwitterung und Erhalt der Gesteine. Berlin Union Deutsche

Verlagsgesellschaft

Garbrecht G (1995) Meisterwerke antiker Hydrotechnik. Einblicke in die

Wissenschaft. B.G. Teubner Verlagsgesellschaft, Stuttgart

Geological map of the Free state Saxony (1992) 1:400,000. Sächsisches

Landesamt für Umwelt und Geologie, 3. Auflage, Freiberg

Glock C (2004) Traglast unbewehrter Beton- und Mauerwerkswände.

Nichtlineares Berechnungsmodell und konsistentes Bemessungskonzept für

schlanke Wände unter Druckbeanspruchung. Institut für Massivbau,

Technische Universität Darmstadt, Heft 9, Darmstadt

Gucci N & Barsotti R (1995) A non-destructive technique for the determination of

mortar load capacity in situ. Materials and Structures, 28, pp 276–283

Herrbach S (1996) Sicherung historischen Mauerwerks am Beispiel des Schlosses

in Schönfeld. Diplomarbeit, Technische Universität Dresden, Institut für

Massivbau

Hilsdorf H (1969) Investigation into the Failure Mechanism of Brick Masonry

Loaded in Axial Compression. Proceedings of International Conference on

Masonry Structural Systems, Houston, Texas, pp 34–41

Huesmann M & Knöfel D (1991) Mörteluntersuchungen von historischem

Mauerwerk am Beispiel der Michaelskirche Fulda. In: Proceedings of the 9th

International Brick/Block Masonry Conference, 13th–16th October 1991, Vol.

1, Berlin, Germany, pp 1406–1411

References 195

zentrischer Druckbeanspruchung: Entwicklung und Anwendung eines Finite-

Elemente-Programmes. Kassel University Press, Kassel

Jackson P (2004) Highways Agency BD on New Masonry Arch Bridges. 27th

February 2004. http://www.smart.salford.ac.uk/February_2004_Meeting.php

Khoo CL & Hendry AW (1972) A Failure Criterion for Brickwork in Axial

Compression. The British Ceramic Research Association, Technical Note, No.

179

Knöfel D & Middendorf B (1991) Chemisch-mineralogische Mörteluntersuchungen

an historischen Ziegelgebäuden in Norddeutschland. In: Proceedings of the

9th International Brick/Block Masonry Conference, 13.–16. October 1991,

Vol. 1, Berlin, Germany, pp 1420–1427

Knöfel D & Schubert P (1991) Zur Beurteilung von Mörtel für historische

Bauwerke. In: Proceedings of the 9th International Brick/Block Masonry

Conference, 13.–16. October 1991, Vol. 1, Berlin, Germany, pp 1412–1419

Lissai, F (1986) Elastisch-plastisches Verhalten von Mauerwerks-Wandscheiben,

die in ihrer Ebene vertikal und horizontal belastet sind. Bautechnik, Heft 1,

pp 7–17

Mann W (1983) Druckfestigkeit von Mauerwerk: Eine statistische Auswertung

von Versuchsergebnissen in geschlossener Darstellung mit Hilfe von

Potenzfunktionen, Mauerwerk-Kalender 1983, pp 687–699

Mann W (1991) Grundlagen der Bemessung von Mauerwerk unter vertikaler

Belastung und Knicken nach dem Entwurf zu Eurocode EC 6 und Vergleich

mit Versuchsergebnissen. Proceedings of the 9th International Brick/Block

Masonry Conference, 13–16 Oct. 1991, Vol 1, Berlin, Germany, pp 1281–1291

Meskouris K, Butenweg C, Mistler M & Kuhlmann W (2004) Seismic Behaviour

of Historic Masonry Buildings. In Proceeding of 7th National Congress on

Mechanics of HSTAM, Chania, Crete, Greece

Müller B (1977) Beiträge zur Geschichte der Natursteinindustrie in der

Sächsischen Oberlausitz. Abhandlungen des Staatlichen Museums für

Mineralogie und Geologie zu Dresden. Band 27, Dresden, pp 111–142

Ohler A (1986) Zur Berechnung der Druckfestigkeit von Mauerwerk unter

Berücksichtigung der mehrachsigen Spannungszustände in Stein und Mörtel.

Bautechnik, Heft 5, pp 163–169

Papayianni I & Stefanidou M (2003) Mortars for intervention in monuments and

historical buildings. Structural Studies, Repairs and Maintenance of Heritage

Architecture VIII, CA Brebbia (Ed), WIT-Press, Southampton, pp 57–64

Pauser A (2005) Ingenieurbüro: Guide to the high-level assessment of masonry

arch bridges. UIC Report (draft)

Peschel A (1984) Natursteine der DDR – Kompendium petrographischer,

petrochemischer, petrophysikalischer und gesteinstechnischer Eigenschaften.

Beilage zur Dissertation (B) an der Bergakademie Freiberg, Januar 1984

Proske D (2003) Ein Beitrag zur Risikobeurteilung alter Brücken unter

Schiffsanprall. Dissertation, Technische Universität Dresden

Purtak F (2001) Tragfähigkeit von schlankem Mauerwerk. Dissertation.

Technische Universität Dresden, Fakultät Architektur

196 4 Masonry Strength

Natursteinbogenbrücken. Bautechnik 8, Heft 8, pp 525–543

Ril 805 (1999) Richtlinie 805: Tragsicherheit bestehender Eisenbahnbrücken

Rustmeier HG (1982) Untersuchungen über Einflüsse auf die Drucktragfähigkeit

von Bruchsteinmauerwerk. Dissertation, Technische Hochschule Darmstadt,

Darmstadt

Sabha A & Schöne I (1994) Untersuchungen zum Tragverhalten von Mauerwerk

aus Elbsandstein. Bautechnik 71, Heft 3, pp 161–166

Sabha A & Weigert A (1996) Einfluss der Steinhöhe auf das Tragverhalten

einschaligen Mauerwerkes. Sonderdruck SFB 315, Erhalten historisch

bedeutsamer Bauwerke, Jahrbuch 1995, Verlag für Architektur und technische

Wissenschaften, Berlin

Schnackers PJH (1973) Mauerwerk und seine Berechnung. Dissertation, TH

Aachen

Schubert P (2004) Eigenschaftswerte von Mauerwerk, Mauersteinen und

Mauermörtel. Mauerwerk-Kalender 1994, Verlag Ernst und Sohn, Berlin, pp

129–141

Schulenberg W (1982) Theoretische Untersuchungen zum Tragverhalten von

zentrisch gedrücktem Mauerwerk aus künstlichen Steinen unter besonderer

Berücksichtigung der Qualität der Lagerfugen, Dissertation, TH Darmstadt

Fachbereich Architektur

Simon E (2002) Schubtragverhalten von Mauerwerk aus großformatigen Steinen.

Dissertation, Technische Universität Darmstadt, Darmstadt

Stein A (1993) Bemessung von Natursteinfassaden. Rudolf Müller Verlag, Köln

Stiglat K (1984) Zur Tragfähigkeit von Mauerwerk aus Sandstein. Bautechnik 2,

pp 54–59 and Bautechnik 3, pp 94–100

Straub H (1992) Die Geschichte der Bauingenieurkunst. Ein Überblick von der

Antike bis in die Neuzeit. Birkhäuser Verlag Basel

UIC-Codex (1995) Empfehlungen für die Bewertung des Tragvermögens

bestehender Gewölbebrücken aus Mauerwerk und Beton. Internationaler

Eisenbahnverband. 1. Ausgabe 1.7.1995

Van der Vlist AA, Bakker MM, Heijnen WJ, Roelofs HJJ & Oosterhoff J (1998)

Bruggen in Nederland 1800–1940. Bruggen van beton, steen en hout.

Nedelandse Bruggen Stichting – Uitgeverij Matrijs

van Hees RPJ, Binda L, Papayianni I & Toumbakari E (2004) Characterisation

and damage analysis of old mortars. Materials and Structures, 37, November

2004, pp 644–648

Verstrynge E, Schueremans L & Van Gemert D (2008) Life time expectancy of

historical masonry structures subjected to creep – a probabilistic approach.

Proceedings of the 6th International Probabilistic Workshop, CA Graubner, H

Schmidt & D Proske (Eds), Darmstadt 2008

Walthelm U (1990) Rissbildungen und Bruchmechanismen in freistehenden

Wänden aus Beton und Mauerwerk. Bautechnik 67, Heft 1, pp 21–26

Warnecke P (1995) Tragverhalten und Konsolidierung von historischem

Natursteinmauerwerk. Dissertation, Technische Universität Braunschweig

Warnecke P (1997) Biegedruckfähigkeit von historischem Mauerwerk – Modelle

und Anwendung. Konsolidierung von historischem Natursteinmauerwerk, 6.

References 197

Baustoffe, Massivbau und Brandschutz Heft 135, Braunschweig, pp 13–28

Warnecke P, Rostasy FS & Budelmann H (1995) Tragverhalten und Konsolidierung

von Wänden und Stützen aus historischem Natursteinmauerwerk.

Mauerwerkskalender 1995, Ernst & Sohn, Berlin, pp 623–685

Wedler B (1997) Einfluss der Füllungsschichtung auf das Tragverhalten des

mehrschaligen Mauerwerks in historischen Bauwerken. Diplomarbeit.

Technische Universität Dresden. Lehrstuhl für Massivbau

Weigert A (1996) Untersuchung zum Einfluss der Steinhöhe auf das

Tragverhalten des einschaligen Mauerwerks aus Elbsandstein. Diplomarbeit.

Lehrstuhl für Massivbau. Technische Universität Dresden

Wenzel F (Ed) (1997) Mauerwerk – Untersuchen und Instandsetzen durch

Injizieren, Vernadeln und Vorspannen. Erhalten historisch bedeutsamer

Bauwerke. Empfehlungen für die Praxis. Sonderforschungsbereich 315,

Universität Karlsruhe

Wisser S & Knöfel D (1988) Untersuchungen an historischen Putz- und

Mauermörteln: Bautenschutz und Bausanierung 11, Nr 5, pp 163–171

5 Investigation Techniques

into things until we see them growing from the beginning.”

5.1 Introduction

The preparation of input data for the numerical modelling of the arch

bridges is a major part of the investigation of arch bridges: a sophisticated

numerical model is without much worth if the quality of the input data is

rather low. Therefore, the observation of the existing structure is of utmost

importance to understand such a structure. Even further, numerical models

and structural observation interact with each other. Whereas for new

structures the choice of the static system is part of the design process, for

existing structures the situation is different: here the static system has to be

identified. Usually in the beginning of such an observation, only very

limited knowledge is accessible. However, with the first numerical models

the required input data and the zones of interest for observation are

identified and investigated. This should improve the numerical model

yielding to further refinement of the interesting parameters. Lourenço

(2001) has described the initial state with the following:

• Important geometrical properties are unknown

• Information about the internal construction is unknown or limited

• The material properties are often unknown or difficult to identify

• The type of construction is unknown

• Possible damages are unknown

• The design basis of the structure is unknown and modern codes are

usually not applicable

• The parameters show significant deviations, either due to the former

construction type or due to the use of natural construction materials

(Goretzky 2000, Franke, Deckelmann and Goretzky 1991 and Kirtschig

1991)

D. Proske, P. van Gelder, Safety of Historical Stone Arch Bridges, DOI 10.1007/978-3-540-77618-5_5,

© Springer-Verlag Berlin Heidelberg 2009

200 5 Investigation Techniques

observation is compulsive. Different technologies can be used for such an

observation and are listed in Table 5-1. Further techniques are summarized

in Schueremans and Van Gemert (2001), Wenzel and Kahle (1993), Kaplan

(1997), Colla (1997), Silman and Ennis (1993), Prieto et al. (2006), and

Orbán et al. (2008). In general, the observation techniques are classified

into destructive, semidestructive and nondestructive identification methods.

Bién and Kamiński (2007) relate certain damages to certain investigation

types (Table 5-2).

(Schueremans et al. 2003, Wenzel and Kahle 1993, Kaplan 1997, Silman and Ennis

1993, Prieto et al. 2006, and Orbán et al. 2008)

Technique Degree of Location General principle and application field

destruction

Historic NDT IS and Historical documents often include

research IL worthwhile information about the con-

struction technology, used materials, and

the geometry. Very often it is useful not

only to investigate files but also to contact

local history association

Visual NDT IS Visual inspection is compelling since it is

inspection cheap and the most efficient nondestructive

test method. It can be extended by monitor-

ing systems or geomarking tools

Photogrammetry NDT IS Photogrammetry can be used to identify

any type of deformations including discon-

tinuities. In recent years, not only

photogrammetry but also other techniques

such as laser scanning and laser interfer-

ence techniques have been applied and

they have yielded extraordinary results

Electric NDT IS This method can be applied to achieve in-

resistivity formation about the overall conditions of

the masonry such as cavities and layering

Radiography NDT IS The application of strong ionizing radiation,

mainly gamma rays, can be used to identify

not only the steel elements in the structure,

but also the cavities and other types of dis-

continuities. However, safety considera-

tions limit the applicability under practical

conditions

5.1 Introduction 201

destruction

Infrared NDT IS Identification of layering of the structure

thermography and further discontinuities

Magnetic NDT IS Identification of steel or iron elements in-

methods side the masonry blocks

Radar NDT IS Radar can give indications about certain

types of discontinuities such as cavities

Mechanical NDT IS Waves are introduced to the material. The

pulse velocity wave velocity gives information about the

integrity and density of the material

Ultrasonic NDT IS Waves are introduced to the material.

Again, information about not only the den-

sity, but also the humidity and discontinui-

ties can be gained. Limited application of

masonry

Vibration tests NDT IS Investigation of the stiffness of the struc-

tural elements

Endoscopy SDT and IS After drilling, endoscopy can be used to

NDT investigate the internal structure. In most

cases, it is combined with video taping

Flat jack SDT IS Determination of the stress-strain relation-

ship, also sometimes used for the identifi-

cation of the maximum compressive

strength

Proof loading NDT IS Testing of load of some structural parts. It

increases the certainty about the applied

numerical models but may cause some

slight damage

Monitoring NDT IS Permanent measurement of certain struc-

tural parameters

Amount of destruction: DT – destructive test; SDT – semi-destructive test;

NDT – non-destructive test

Location of test: IS – in situ; IL – in labo.

202 5 Investigation Techniques

(MDT) for the investigation of damages to masonry bridges (Bién and Kamiński

2007)

Test Degradation mechanism Damage type

Loos of material

contamination

Displacement

discontinuity

deformation

destruction

Basic methods

Visual inspections

Direct geometric measurements

Sclerometric test

NDT Acoustic and stress wave methods

Acoustic emission measurement

Impact echo test

Parallel seismic method

Ultrasonic echo test

Electrical and electromagnetic methods

Electrical conductivity measurement

Ground penetrating radar

Thermal heat transfer methods

Pulse-phase thermography

Transient thermography

Proof load tests

Dynamic test

Static tests

Boroscopy

Flat-jack test

MDT Pull out test

Specimen test – chemical

Specimen test – mechanical

the material properties. Such material properties are often measured on test

specimens. The specimens have to be separated from the original struc-

tures. A wide literature for the extraction of test specimens from historical

masonry propagating many different techniques can be found.

For example, Stiglat (1984) recommends the extraction of unnecessary

stones from the structure. However, under realistic conditions, it is difficult to

5.2 Destructive Tests 203

the functionality of the structure. Furthermore, only visible alterations of a his-

torical structure are often prohibited due to conservation regulations of

monuments and historic buildings (Budelmann 1997, Wenzel 1997a, b).

Therefore, in many cases, drilling of cores has to be carried out. Although

this type of material extraction is often criticized due to uncertainties about

the process (Stiglat 1984 and Berndt and Schöne 1990), the technology is

simple to apply even under difficult conditions, for example under water.

Furthermore, the drillings not only allow the extraction of material but also

provide geometrical data about the internal structure. Additionally, the visual

disturbance of the structure due to drillings is rather limited since only the di-

ameter of the drilling machine has to be substituted on the structure surface.

Besides the material excavation, drilling also provides data from the

process alone. For example, the volume of cooling water during the drill-

ing process permits conclusions about the pore volume in the masonry. Fi-

nally, drilling can be combined with endoscopy, which allows visual views

inside the structure. After the material has been used for material testing,

the data can be compared with the drilling protocols.

The diameter of the core drilling depends on different conditions. In

general, from the material investigation point of view, drilling diameters

should be rather great to achieve characteristic material property data. For

example, Stiglat (1984) recommends a minimum diameter of 20 cm for

natural stone masonry walls. However, such big diameters are often not

applicable due to visual disturbance of the structure, increased breaking

rate of the cores to the high friction, and simply drilling costs or drilling

conditions. Therefore, under practical conditions, often diameters in the

range of 10–15 cm are used. Under specific conditions, diameters of 5 cm

are also used; however, the measured material data may then be of re-

stricted use. Wenzel (1997b) recommends minimum diameters of 3 cm for

brick stones and 5 cm for natural stones.

After the drilling process, the drilling cores are visually observed (Fig. 5-1).

The observation is used to create a drilling profile, for example, identifying

layer thickness, rough material estimation, and identification of hollow

sections. The layers are then classified according to the size and number of

material pieces (lumpy, small sized) (DIN 4022 1987). Based on the classi-

fication, possible test specimens for material testing can be marked on the

cores. The specimens are then sawed out. However, sawing requires cool-

ing water and the water can influence some material properties. It can even

prevent the production of test specimens. The material can also be used for

chemical, petrographic (Fig. 5-2), spectrographic, and microscopic investi-

gations (Fig. 5-3).

A description of drilling investigations at a historical arch bridge can be

found in Aoki et al. (2004) or Proske (2003).

204 5 Investigation Techniques

Fig. 5-1. Example of drilling cores. See also the marking of test specimen for

material property tests

Fig. 5-2. Different sandstone varieties taken from one historical arch bridge

5.4 Non-destructive Test Methods 205

also called minor-destructive test methods, can be distinguished into

• Pull-out tests

• Pull-off tests

• Penetration tests (Windsor-Probe, Schmidt hammer)

For further information, refer Corps of Engineers (2002).

(2002), Orbán et al. (2008), and Bungey (1997). Nondestructive test methods

can provide the following information about arch bridges (Forde 1996):

• Type and construction of the springing

• Thickness of the arch

• Type of backfill and existence of vaults inside the backfill

• Density of the backfill

Astudillo (1996) presents nondestructive investigation methods for the

examination of bridges. Colla et al. (1997) report about nondestructive techni-

ques on stone masonry bridges. Bensalem et al. (1998) use nondestructive

investigation methods for the estimation of the safety factor decrease. Aoki

et al. (2004) and Binda and Saisi (2001) have used nondestructive tech-

niques for the investigation of masonry. With regard to masonry, the Euro-

pean research project ONSITEMASONRY should be mentioned, which

focused on development of nondestructive and semidestructive investiga-

tion techniques for masonry (Wendrich et al. 2004, Maierhofer et al. 2003,

and Köpp et al. 2005). During the sustainable bridge project of the EU,

nondestructive tests for the assessment of bridges were also evaluated

(Helmerich and Niederleithinger 2006 and Niederleithinger et al. 2006).

Recently, Orbán et al. (2008) showed some application for historical arch

bridges.

5.4.1 Ultrasound

stone structures has found wide application. As examples, the works by

Schubert et al. (2002) and Müller and Garke (2005) should be mentioned.

206 5 Investigation Techniques

The general idea is simple: the denser a material is or better the material

joints are, the better mechanical waves can spread out in the material. The

application of ultrasound uses this effect. For the investigation, sound waves

in a frequency range between 46 and 350 kHz are introduced in the material

investigated. The waves are then registered at a different location. The time

difference between the sending and the reception of the waves is measured

and the velocity of the ultrasonic waves is then computed. The sound veloc-

ity depends on certain material parameters, such as the material structure and

the amount of pores. Furthermore, humidity or watering conditions of the

material can influence the ultrasonic velocity. Due to these effects, a drop in

the wave velocity may have different causes, which under practical condi-

tions may restrict the application of ultrasonic techniques.

5.4.2 Impact-echo

sonic waves. In contrast, here a hammer is used to apply shock waves to

the material. The echo is again measured with receivers. If material defects

occur in the structural material, then the echo-impulses are decreased in

comparison to homogenous materials (Leaird 1984).

Values for average ultrasonic velocities are given in Table 5-3.

According to Forde (1996), the frequency range lies between 1 kHz and

300 Hz. Forde (1996) has already used the impact-echo technique for in-

vestigations of the springings and abutments of historical arch bridges.

Forde (1996)

Material Average ultrasonic velocity in m/s

High quality brick masonry 3,100

Low quality brick masonry 2,500–2,700

Structural concrete >4,500

Granite masonry piers 3,300–3,500

Red sandstone piers 1,970

Yellow sandstone piers 2,040

White sandstone piers 1,700

Steel bar 5,100

Steel body 6,100

Dry sandy ground 200–300

Dry sandy clay 400–600

Water saturated clay 1,300–2,400

Water 1,430–1,680

Limestone and dolomite 4,000–6,000

5.4 Non-destructive Test Methods 207

5.4.3 Radar

The application of radar technology for masonry and masonry arch bridges

is strongly related to the development of ground penetration radar. This ra-

dar has been widely applied for the investigation of ground structures.

Based on the experience gained there, the technique has also been applied

for archeological investigations, for the investigation of structures in the

ground, and for investigation of buried vaults and foundation rests. Exam-

ples are given in Kahle and Illich (1992), Illich (1999), BAM (2006a,

b), Wiggenhauser and Maierhofer (2002), Cameron et al. (2008), and

Wenzel (1997b).

During a radar investigation, waves (now radar waves) are also entered

into the structure. If the waves hit irregularities such as separated surfaces,

change in the salt content and humidity, hollow cavities, or metal elements

inside the element, then the waves are reflected. Since the sender and re-

ceiver are usually assembled jointly into one case, the intensity and the

running time of the waves can be used to compute the depth of the reflec-

tion zone.

The radar case is moved slowly over the surface of the structural ele-

ment to achieve not only results at some certain points, but also to cover

surfaces. Using the depth information, three-dimensional maps can be pre-

pared.

Very often 1 GHz antennas are used. Such antennas reach a penetration

depth of 1.0–1.5 m. The resolution depends on the frequency and reaches 3

cm for 1 GHz. For masonry arch bridges, lower frequencies in the range of

100 MHz are often used according to Forde (1996). Here, radar investiga-

tions are mainly used to identify hollow sections inside piers or walls, to

find cracks, and to estimate the salt content and the water content (Forde

1996). Clark et al. (2003a, b) describe the application of radar and infrared

for the humidity investigation.

To give an impression about the quality of radar and ultrasonic investi-

gations, Table 5-4 lists the results of such an investigation. The table also

includes the achieved maximum flexural bending stress of the granite

stone material investigated. The measured wave velocities and found ir-

regularities are already transferred into qualitative statements about the

flexural bending strength. All methods find the one weak stone (Nr. 1),

however for the other stones the results are inconsistent.

208 5 Investigation Techniques

Table 5-4. Comparison of the flexural bending strength of granite stone material

and the results from radar and ultrasonic investigation

Test Qualitative statement Qualitative state- Qualitative state- Maximum

number about the strength ment about the ment about the flexural bend-

according to strength according strength according ing strength

ultrasonic test 1 to ultrasonic test 2 to radar test in MPa

1 Low Low Low 4.23

2 Average to good – Good to very good 9.66

3 Average to good Good to very good Good to very good 12.70

4 Good to very good Low to average Average to good 9.53

5 Good to very good – Low strength 10.21

5.4.4 Tomography

Bridge Le Pont-Neuf in Paris. At first, 20 single-spot tomographical meas-

urements were taken on the pregrouted pier. Based on that data, a spatial

picture of the velocity distribution of ultrasonic waves inside the pier was

constructed. Then the velocity was translated into the density of the mate-

rial. The results showed a very inhomogeneous limestone inside the pier.

In the next step, the pier was grouted and again investigated tomographi-

cally. As the result of the grouting, the more homogeneous density field

was found. Computer tomography as presented by Schulze and Hampel

(2004) may only be used for some small structural elements.

5.4.5 Thermography

masonry walls (Orbán et al. 2008). However, the technique is used more

for buildings than for bridges.

Conductivity tests can be used for salinity and stone thickness investiga-

tions (Forde 1996 and Helmerich et al. 2008).

Besides the introduced techniques, there exists finally the possibility to apply

a load test on the arch bridge on site. Basic works for such experimental

5.4 Non-destructive Test Methods 209

some examples of application for historical arch bridges should be men-

tioned here.

Mildner (1996) describes load tests and deformation measurements of

the Schrote Bridge and the Anna-Ebert Bridge in Magdeburg. Further tests

by Milder can be found in Mildner and Mildner (2001). Vockrodt and

Schwesinger (2002) have also carried out experimental load-bearing tests

on historical arch bridges. Steffens (2001), Gutermann and Steffens

(2005), Burkert and Steffens (2008), and Gutermann (2002) give recom-

mendations about the application of in situ load tests and show examples.

Fanning and Boothby (2003) also report about load tests on arch bridges.

Bolle (2005) describes the permanent observation of a viaduct. Slowik et

al. (2005) and Slowik (2004) also report on load testing on vault bridges

and the development of the numerical model based on the data. Domède

and Sellier (2008) have also carried out measurements and have used it to

create an FEM model. Hughes and Pritchard (1998) have published on in

situ measurement of masonry arch bridges. Armstrong et al. (1995a) have

carried out dynamic measurements on arch bridges and have continued with

modal analyses (Armstrong et al. 1995b). Measurement under load was also

done by Bién et al. (2008), Prader et al. (2008), and Rücker et al. (2006).

Measurement of fill pressure was done by Ponniah and Prentice (1999).

Fibre optic sensors can also be applied for strain measurements on arch

bridges (Inaudi and Glisic 2008).

During the observation of arch bridges under load, certain different meth-

ods can be applied to receive the deformations data. Photogrammetry is

one of the contact-free techniques (Hampel and Maas 2003). Albert and

Seyler (2004), for example, have used photogrammetry for the investiga-

tion of arch bridge deformations.

A second technique is laser scanning. Laser scanning is not only used to

scan snow covers in the mountains now (Prokop 2007) but also widely ap-

plied since a few years for the capture of structural geometries (Ehmann

2000, Mönicke 2003).

210 5 Investigation Techniques

References

Bildsequenzen, 14th International Conference on Engineering Surveying

Zürich, 15.–19. März 2004

Aoki T, Komiyama D, Sabia D & Rivella R (2004) Theoretical and Experimental

Dynamic Analysis of Rakanji Stone Arch Bridge, Honyabakei, Oita, Japan,

Proceedings of 7th International Conference on Motion and Vibration Control

MOvIC’04, St. Louis, 8.-11. August, 2004, pp 9 (CD-ROM)

Aoki T, Komiyama T, Tanigawa Y, Hatanaka S, Yuasa N, Hamasaki H, Chiorino

MA & Roccati R (2004) Non-destructive Testing of the Sanctuary of Vico-

forte, Proceedings of 13th International Brick and Block Masonry Confer-

ence, Amsterdam, July 4–7, 2004, Vol. 4, pp 1109–1118

Armstrong DM, Sibbald A & Forde MC (1995a) Integrity assessment of masonry

arch bridges using the dynamic stiffness technique. NDT & E International,

28 (6), pp 367–375

Armstrong DM, Sibbald A, Fairfield CA & Forde MC (1995b) Modal analysis for

masonry arch bridge spandrel wall separation identification. NDT & E

International, 28 (6), pp 377–386

Astudillo R (1996) Use of nondestructive methods for bridge auscultation in

Spain. Recent Advances in Bridge Engineering. Evaluation, Management and

Repair. Proceedings of the US-Europe Workshop on Bridge Engineering, or-

ganized by the Technical University of Catalonia and the Iowa State Univer-

sity, Barcelona, 15–17 July 1996, International Center for Numerical Methods

in Engineering CIMNE, pp 202–213

BAM (2006a) http://www.bam.de/service/publikationen/zfp_kompendium/

geraete/g050/g050_L.html 2006

BAM (2006b) http://www.ndt.net/article/dgzfp01/papers/p45/fig4_6.htm

Bensalem A, Ali-Ahmed H, Fairfield CA & Sibbald A (1998) NDT as a tool for

detection of gradual safety factor deterioration in loaded arches. Arch Bridges:

History, Analysis, Assessment, Maintenance and Repair. Proceedings of the

Second International Arch Bridge Conference, A. Sinopoli (Ed), Venice 6.–9.

October 1998, Balkema, Rotterdam, pp 271–277

Berndt E & Schöne I (1990) Tragverhalten von Natursteinmauerwerk aus

Elbesandstein. Sonderforschungsbereich 315, Universität Karlsruhe, Jahrbuch

1990

Bień J & Kamiński T (2007) Damages to masonry arch bridges – proposal for

terminology unification. Proceedings of the Arch 07 – 5th International Con-

ference on Arch Bridges, PB Lourenço, DV Oliveira & A Portela (Eds), 12–

14 September 2007, Madeira, pp 341–348

Bién J, Kamiński T & Trela C (2008) Numerical analysis of old masonry bridges

supported by field tests. Bridge Maintenance, Safety, Management, Health

Monitoring and Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis

Group, London, pp 2384–2391

Binda L & Saisi A (2001) Non destructive testing applied to historic buildings:

The case of some Sicilian Churches. Historical Constructions, PB Lourenço &

P Roca (Eds), University of Minho, Guimarães, pp 29–46

References 211

Lösung zur Datenanalyse und Zustandsbewertung. 3. Symposium

Experimentelle Untersuchung von Baukonstruktionen, Technische Universität

Dresden

Budelmann H (1997) Zustandserfassung und -beurteilung historischer

Mauerwerke. Konsolidierung von historischem Natursteinmauerwerk, 6–7.

November 1997, Technische Universität Braunschweig, Institut für Baustoffe,

Massivbau und Brandschutz Heft 135, Braunschweig, pp 53–66

Bungey, J (1997) Non-destructive testing in civil engineering. In: Proceedings of

the International Conference in Civil Engineering, JH Bungey (Ed), Liver-

pool, 08.–11. April 1997, The British Institute of NDT, Northampton 1997,

Vol. 1 and 2

Burkert T & Steffens K (2008) Instandsetzung und Ertüchtigung von Mauerwerk,

Teil 7: Experimentelle Bestimmung der Tragfähigkeit von Mauerwerk,

Belastungsversuche an Mauerwerksbauten in situ, Mauerwerk-Kalender 2009,

W Jäger (Ed), Ernst und Sohn, Berlin

Cameron J, Demoulin C, Diamanti N, Giannopoulos A & Forde MC (2008) As-

pects of GPR testing of masonry arch bridges. Structural Faults + Repair-

2008: 12th International Conference, MC Forde (Ed), Edinburgh, 10th–12th

June 2008

Clark MR, Parsons P, Hull T & Forde MC (2003a) Case study of radar laboratory

work on the masonry arch bridge. Extending the Life of Bridges, Concrete

and Composites Buildings, Masonry and Civil Structures, MC Forde (Ed),

The Commonwealth Institute, London, 1st–3rd July 2003

Clark MR, Parsons P, Hull T & Forde MC (2003b) Laboratory study of infrared

thermography on masonry arch bridges. Extending the Life of Bridges, Con-

crete and Composites Buildings, Masonry and Civil Structures, MC Forde

(Ed), The Commonwealth Institute, London, 1st–3rd July 2003

Colla C (1997) NDT of masonry arch bridges, PhD Thesis, The University of

Edinburgh, Department of Civil and Environmental Engineering, Edinburgh

Colla C, Das, PC, McCann D & Forde MC (1997) Sonic, electromagnetic and im-

pulse radar investigation of stone masonry bridges. NDT&E International, 30

(4), pp 249–254

Corps of Engineers (2002). Nondestructive Evaluation Techniques, Chapter 10,

TM 5-809-3/NAVFAC DM-2.9/AFM 88-3

Cote P (1996) NDT applied to concrete and masonry bridges. Recent Advances in

Bridge Engineering. Evaluation, Management and Repair. Proceedings of the

US-Europe Workshop on Bridge Engineering, JR Casas, FW Klaiber & AR

Marí (Eds), Organized by the Technical University of Catalonia and the Iowa

State University, Barcelona, 15–17 July 1996, International Center for Nu-

merical Methods in Engineering CIMNE, pp 192–201

DIN 4022 (1987) Part 1–3: Baugrund und Grundwasser; Benennen und

Beschreiben von Boden und Fels; Schichtenverzeichnis für Bohrungen

mit/ohne durchgehende Gewinnung von gekernten Proben im Boden und Fels

(Fest- und Lockergestein), September 1987

212 5 Investigation Techniques

Structural faults + Repair-2008: 12th International Conference, MC Forde

(Ed), 10th–12th June 2008, Edinburgh, UK

Ehmann D (2000) Untersuchung und Erprobung der zielmarkierungsfreien

Deformationsmessung an Bauwerken im Nahbereich. Diplomarbeit, HfT

Stuttgart, Fb. Vermessung und Geoinformatik

Fanning PJ & Boothby TE (2003) Experimentally-based assessment of masonry

arch bridges. Bridge Engineering, 156 (3), September 2003, pp 109–116

Forde MC (1996) Non-destructive evaluation of bridges: research in progress. Re-

cent Advances in Bridge Engineering. Evaluation, management and repair.

Proceedings of the US-Europe Workshop on Bridge Engineering, JR Casas,

FW Klaiber, AR Marí (Eds), Organized by the Technical University of Cata-

lonia and the Iowa State University, Barcelona, 15–17 July 1996, Interna-

tional Center for Numerical Methods in Engineering CIMNE, pp 127–152

Franke I, Deckelmann G & Goretzky W (1991) Einfluss der Streubreite der

Materialeigenschaften auf die Tragfähigkeit von Mauerwerk – Theoretische

und Experimentelle Untersuchungen. In: Proceedings of the 9th International

Brick/Block Masonry Conference, 13.–16. October 1991, Vol. 1, Berlin,

Germany, pp 180–187

Goretzky W (2000) Tragfähigkeit druckbeanspruchten Mauerwerks aus

festigkeits- und verformungsstreuenden Mauersteinen und -mörteln.

Dissertation GCA Verlag Herdecke

Gutermann M & Steffens K (2005) Belastungsversuche zur Bausubstanzerhaltung.

3. Symposium Experimentelle Untersuchung von Baukonstruktionen, Techni-

sche Universität Dresden

Gutermann M (2002) Ein Beitrag zur experimentell gestützten Tragsicherheitsbe-

wertung von Massivbrücken. Dissertation, Technische Universität Dresden

Hampel U & Maas H-G (2003) Application of digital photogrammetry for meas-

uring deformation and cracks during load tests in civil engineering material

testing. In: 6th Conference on Optical 3-D Measurement Techniques. Prof. A.

Grün, Institute of Geodesy and Photogrammetry, Swiss Federal Institute of

Technology, Zürich, 22–25.9.2003, Vol. II, pp 80–88

Helmerich R & Niederleithinger E (2006) NDT-toolbox for the Inspection of

Railway Bridges. Sustainable Bridges Project – Assessment for Future Traffic

Demands and Longer Lives.

Helmerich R, Niederleithinger E, Trela C, Bien J & Bernardini G (2008) Complex

multi-tool inspection of a masonry arch bridge using non-destructive testing.

Bridge Maintenance, Safety, Management, Health Monitoring and Informat-

ics, H Koh & DM Frangopol (Eds), Taylor & Francis Group, London, pp

3716–3723

Hughes TG & Pritchard R (1998) In situ measurement of dead and live load

stresses in a masonry arch. Engineering Structures, 20 (1–2), pp 5–13

Illich B (1999) Zerstörungsfreie Zustandserfassung an Mauer- und Betonbauteilen

mit Radar. DGZfP-Berichtsband 66-CD, Vortrag 9, 21.–22. Januar 1999 in

der Neuen Messe München. pp 89–93

References 213

Inaudi D & Glisic B (2008) Overview of 40 bridge monitoring projects using fiber

optic sensors. Bridge Maintenance, Safety, Management, Health Monitoring

and Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis Group,

London, pp 2514–2521

Kahle M & Illich B (1992) Erkundung historischen Mauerwerks mit Radar. 3.

Internationales Kolloquium für Werkstoffwissenschaften und Bausanierung,

Ehningen, pp 379–393

Kaplan ME (1997) Non-destructive evaluation techniques for masonry

construction (Masonry), U.S. Dept. of the Interior, National Park Service,

Cultural Resources, Heritage Preservation Services

Kirtschig K (1991) Zur Beschreibung der Mauerwerksdruckfestigkeit über die

mittlere oder charakteristische Steindruckfestigkeit. In: Proceedings of the 9th

International Brick/Block Masonry Conference, 13.–16. October 1991, Vol. 1,

Berlin, Germany, pp 196–201

Köpp C, Maierhofer C & Wendrich A (2005) Entwicklung zerstörungs-freier und

zerstörungsarmer Prüfverfahren zur Untersuchung historischen Mauerwerks.

Mauerwerk 9, Heft 3, pp 102–107

Leaird JD (1984) A report on the pulsed acoustic emission technique applied to

masonry, Journal of Acoustic Emission, 3 (4), pp 204–210

Lourenço PB (2001) Analysis of historical constructions: From thrust-lines to ad-

vance simulations. Historical Constructions, PB Lourenço & P Roca (Eds),

University of Minho, Guimarães, pp 91–116

Maierhofer C, Wendrich A & Köpp C (2003) ONSITEFORMASONRY – A

European Research Project: On-site investigation techniques for the structural

evaluation of historic masonry, DGZfP (Ed) International Symposium Non-

Destructive Testing in Civil Engineering (NDT-CE) in Berlin, Germany, Sep-

tember 16–19, 2003, Proceedings on BB 85-CD, V70, Berlin

Mildner K & Mildner A (2001) Bauwerksmessungen vor und nach der

Instandsetzung einer Eisenbahn-Gewölbebrücke und statische Modellbildung.

Ehrenkolloquium Prof. Dr.-Ing. Möller zum 60. Geburtstag, Lehrstuhl für

Statik und Dynamik der Tragwerke, Technische Universität Dresden, 2001,

Dresden

Mildner K (1996) Tragfähigkeitsermittlung von Gewölbebrücken auf der

Grundlage von Bauwerksmessungen. 6. Dresdner Brückenbausymposium, 14.

März 1996, Fakultät Bauingenieurwesen, Technische Universität Dresden, pp

149–162

Mönicke H-J (2003) Profil- und flächenhafte Objektvermessung mit Tachymeter

oder 3D-Laser-Scanner. In: Ingenieurblatt für Baden-Württemberg, Heft

4/2003, pp 102–106

Müller K-H & Garke R (2005) Auswertung von Ultraschallmessungen unter

Berücksichtigung von Vorinformationen. Revitalisierung von Bauwerken.

Schriften der Bauhaus-Universität, Heft 117, pp 33–40

Niederleithinger R et al. (2006) Guideline for Condition Assessment and

Inspection of Railway Bridges. Sustainable Bridges Project – Assessment for

Future Traffic Demands and Longer Lives

O’Connor A, Pedersen C, Enevoldsen I, Gustavsson L & Hammarbäck J (2008)

Probability based assessment of a large riveted truss railway bridge. Bridge

214 5 Investigation Techniques

Koh & DM Frangopol (Eds), Taylor & Francis Group, London, pp 1812–1819

Orbán Z, Yakovlev G & Pervushin G (2008) Non-Destructive Testing of masonry

arch bridges – an overview. Bautechnik (85), Heft 10, pp 711–717

Ponniah DA & Prentice DJ (1999) Long term monitoring of fill pressure in a new

brickwork arch bridge. Construction and Building Materials 13, pp 159–167

Prader J, Weidner JW, Hassanain H, Moon F, Aktan E, Jalinoos F, Buchanan B &

Ghasemi H (2008) Load testing and analysis of bridges missing critical

documentation. Bridge Maintenance, Safety, Management, Health Monitoring

and Informatics, H Koh & DM Frangopol (Eds), Taylor & Francis Group,

London, pp 1877–1885

Prieto F, Martin M & Viola M (2006) Application of laser vibromety analysis to

damage detection in masonry structures. Heritage, Weathering and Conserva-

tion, R Fort, M de Buergo, M Gomez-Heras & C Vazquez-Calvo (Eds),

Taylor & Francis Group, London, Volume 2, pp 627–632

Prokop A (2007) The application of terrestrial laser scanning for snow and ava-

lanche research. Universität für Bodenkultur Wien

Proske D (2003) Ein Beitrag zur Risikobeurteilung alter Brücken unter

Schiffsanprall. Dissertation, Technische Universität Dresden

Rücker W, Hille F & Rohrmann R (2006) Guideline for the Assessment of Exist-

ing Structures. SAMCO Final Report 2006

Schubert F, Köhler B & Lausch R (2002): Zerstörungsfreie Ultraschallprüfung von

Betonbauteilen – Impulsecho und Impactechomethode in Theorie und Praxis. 2.

Symposium Experimentelle Untersuchungen von Baukonstruktionen –

Ehrenkolloquium Prof. Dr.-Ing. habil. Heinz Opitz. Heft 17 der Schriftenreihe

des Institutes für Tragwerke und Baustoffe, Fakultät Bauingenieurwesen,

Technische Universität Dresden, pp 25–34

Schueremans L & Van Gemert D (2001) Assessment of existing masonry struc-

tures using probabilistic methods – state of the art and new approaches.

STRUMAS 2001 18.–20. April 2001, Fifth International Symposium on

Computer Methods in Structural Masonry

Schueremans L, Van Rickstall F, Ignoul S, Brosens K, Van Balen K & Van

Gemert D (2003) Continuous Assessment of Historic Structures – A State of

the Art of applied Research and Practice in Belgium. KULeuven, Department

of Civil Engineering

Schulze M & Hampel U (2004) Erfassung und Analyse von Prüfkörpern in der

Materialprüfung durch Computertomographie. In: Publikationen der

Deutschen Gesellschaft für Photogrammetrie, Fernerkundung und

Geoinformation. Band 13, Halle (Saale), Germany, 2004, pp 317–324

Silman R & Ennis, M (1993) Non-destructive evaluation to document historic

Structures, International Association for Bridge and Structural Engineering

(IABSE), Zurich, pp 195–203

Slowik V (2004) Combining experimental and numerical methods for the safety

evaluation of existing concrete structures, Fracture Mechanics of Concrete

Structures, VC Li, CKY Leung, KJ Willam & SL Billington (Ed), Proceed-

ings of Framcos-5, April 12–16, 2004, Vail, USA, IA-FraMCoS, Volume 2,

pp 701–707

References 215

experimenteller und rechnerischer Methoden bei der Tragsicherheitsbewertung

bestehender Massivbauwerke, 3. Symposium Experimentelle Untersuchungen

von Betonkonstruktionen, Technische Universität Dresden, Institut für

Massivbau, 23. Juni 2005, Tagungsband, pp 157–168

Steffens K (2001) Experimentelle Tragsicherheitsbewertung von Bauwerken.

Grundlagen und Anwendungsbeispiele. Ernst und Sohn Verlag

Stiglat K (1984) Zur Tragfähigkeit von Mauerwerk aus Sandstein. Bautechnik

2/1984, pp 54–59 and Bautechnik 3/1984, pp 94–100

VanMarcke E (1997) Quantum Origins of Cosmic Structure. Balkema, Rotterdam

Vockrodt H-J & Schwesinger P (2002) Experimentelle Tragfähigkeitsanalyse

einer historischen Bogenbrücke in Erfurt, Bautechnik 79, Heft 6, pp 355–367

Wendrich A, Maierhofer C, Kuritz M & Köpp C (2004) Aufbau eines historischen

Mauerwerkskörpers zur Bewertung von zerstörungsfreien Verfahren.

Jahrestagung 2004 Zerstörungsfreie Materialprüfung. Kurzfassungen der

Vorträge und Plakatbeiträge. Salzburg, 17.–19. Mai 2004. Deutsche,

Österreichische und Schweizerische Gesellschaft für Zerstörungsfreie Prüfung

Wenzel F & Kahle M (1993) Indirect methods of investigation for evaluating

historic masonry, International Association for Bridge and Structural

Engineering (IABSE), Zürich 1993, pp 75–90

Wenzel F (1997a) Behutsame Vorgehensweise bei der Mauerwerksinstandsetzung.

Tagung Konsolidierung von historischem Natursteinmauerwerk, 6.–7.

November 1997, Technische Universität Braunschweig, Institut für Baustoffe,

Massivbau und Brandschutz Heft 135, Braunschweig, pp 1–11

Wenzel F (1997b) (Ed) Mauerwerk – Untersuchen und Instandsetzen durch

Injizieren, Vernadeln und Vorspannen. Erhalten historisch bedeutsamer

Bauwerke. Empfehlungen für die Praxis. Sonderforschungsbereich Karlsruhe.

Universität Karlsruhe

Wiggenhauser H & Maierhofer C (2002) Zerstörungsfreie Bauwerksuntersuchun-

gen mit Radar. 2. Symposium Experimentelle Untersuchungen von Baukon-

struktionen – Ehrenkolloquium Prof. Dr.-Ing. habil. Heinz Opitz. Heft 17 der

Schriftenreihe des Institutes für Tragwerke und Baustoffe, Fakultät Bauinge-

nieurwesen, Technische Universität Dresden, pp 71–78

6 Damages and Repair

6.1 Introduction

er it as an illness, but as a normal process. It is defined as a decrease of the

capability of the organism to cope with the requirements of the environ-

ment with increasing age. Aging can also be related to decrease of reserves

(Strasser 2006).

Such a process can also be found in the case of technical products. The

engineer has to consider the aging of structures during the design process,

if the safety requirements are to be fulfilled over the entire lifetime of the

structure. Design life spans for certain structures are shown in Table 6-1.

So, the structure should function well, not only if it is new but also at the

end of its lifetime. This is sometimes called “graceful degradation”. We do

not only desire it for humans, it is also a desired criterion for technical

products.

Table 6-1. Design work time of concrete structures according to the Eurocode 1

Design life in year Example

1 ... 10 Temporary structures

10 ... 25 Replaceable structural parts, e.g. gantry girders and bearings

15 ... 30 Agricultural used structures

50 Buildings and other common used structures

100 Monument building structures, bridges and other structures

D. Proske, P. van Gelder, Safety of Historical Stone Arch Bridges, DOI 10.1007/978-3-540-77618-5_6,

© Springer-Verlag Berlin Heidelberg 2009

218 6 Damages and Repair

6.2.1 Overview

identified such as carbonation, chloride attack, sulphate attack, and Frost-

Thaw (DIN 1045-1 2001). Also for historical masonry arch bridges, such

durability loads have been identified. They usually yield to changes of

properties of the structure and furthermore to damages. A definition of the

term “damage” can be found in Chapter 7. Such changes and damages can

be found on many historical arch bridges. However, that is not mainly be-

cause these types of structure have been designed so weak, but because

many of these bridges are quite old.

A rough list of damages was given by Bién and Kamiński (2004). They

list the following damages:

• Incompatible deformations (deformations which yield to changes of the

initial geometry)

• Destruction of material caused either by chemical or by physical proc-

esses

• Material discontinuities (cracks)

• Loss of material (falling stones)

• Damage on auxiliary elements (damaged sealing)

• Deformation damages (deformation on the structure that does not yield

to a change of the initial geometry, for example sliding spandrel walls)

• Contamination (natural cover, besmirch)

Typical damage patterns for arch bridges were shown by Angeles-

Yáñez and Alonso (1996), as shown in Fig. 6-1. A classification of damage

patterns for historical stone arch bridges of the European railway organiza-

tions was also given by Orbán (2004) and is summarized in Table 6-2. The

latest catalogue of damages was presented by Bień and Kamiński (2007) in

relation to degradation processes and damages (Table 6-3). Mildner (1996)

has also mentioned typical damages on masonry arch bridges.

6.2 Damages on Historical Arch Bridges 219

Fig. 6-1. Most frequent damages found on arch bridges according to Angeles-Yáñez

and Alonso (1996)

Table 6-2. Types of damages at arch bridges of railways organizations and their

frequency according to Orbán (2004)

Nr. Type of damage1 Frequency2

1 Damage at sealing3 2.1

2 Deterioration of material 2.4

3 Separation and movement of wing wall 3.0

4 Separation and movement of spandrel wall 3.5

5 Damages at piers, foundation and skewback 4.0

6 Geometrical problems with the structure 4.0

7 Other problems4 4.0

8 Cracks in arch caused by settlement 4.2

9 Damages at the road crossing construction 4.3

10 Damages caused by overload 4.3

220 6 Damages and Repair

11 Deformation 4.4

12 Cracks in arch caused by overload 4.5

13 Damages at the parapet caused by single loads 4.6

1

In general, in many cases, the cause of the damage cannot be identified.

2

Calculated as mean value based on information provided by the different railway

organizations. The numbers represent the following:

1 = Very frequent = about 50% of all bridges

2 = Frequent = about 25% of all bridges

3 = Occasional = about 10% of all bridges

4 = Rare = about 5% of all bridges

5 = Exceptional = less than 5% of all bridges

3

Many historical arch bridges were built without sealing. But of course, damages

caused by water can be found there. These bridges have therefore been added to

this statistic.

4

Other problems include damages caused by plants, damages caused by earth-

quakes, impacts and wrong maintenance.

(Bién and Kamiński 2007)

Degradation mechanism Damage type

Loos of material

contamination

displacement

discontinuity

deformation

destruction

Physical

Effects of high temperature

Fatigue

Freeze-Thaw

Change of foundation conditions

Overloading

Shrinkage

Water penetration

Chemical

Carbonation

Crystallization

Leaching

Salt and acid actions

Biological

Accumulation of contamination

Living organisms activities

6.2 Damages on Historical Arch Bridges 221

Recent failures of arch bridges were often related to accidental loads. For

example, the historic Pöppelmann arch bridge in Grimma was heavily

damaged during the 2002 flooding of the river Mulde (Fig. 6-2). The

bridge had to be blasted afterwards since reconstruction using the remain-

ing parts was not possible (Curbach et al. 2003a). Another example was

the failure of the arch bridge in Benairbeig over the Rio Girona in Spain

due to a flashflood called Gota Fría in October 2007 (Meyer 2007). The

failure was actually filmed because television was reporting about the

flashflood onsite. The movies are visible on YouTube. A further example

was the flood-related failure of a farm track and public footpath masonry

arch bridge over the river Devon in 2007 (Bottesford Living History 2007).

Ural et al. (2008) also mentions the failure of Turkish arch bridges by floods.

Flooding and ice loads often caused failures of historical arch bridges.

Drdácký and Slízková (2007) report on the repeated damages and partial

failures of the historical Charles Bridge in Prague caused by flooding in

1359, 1367, 1370, 1373, 1374, 1432, 1496, 1503, 1655, 1784, 1890, and in

2002.

Furthermore, not only are historical bridges exposed to flooding but also

to all gravity-driven mass movements such as debris flows, rock falls, and

avalanches.

Examples of debris flow impacts against historical arch bridges can be

found in Proske (2009). For example, in Log Pod Mangartom in Slovenia,

a huge debris flow killed several people, and destroyed houses and also

one historical arch bridge. The arch bridge over the Lattenbach in Austria

222 6 Damages and Repair

in September 2008. Ural et al. (2008) mention fluvial mass transport in-

cluding dead wood as cause of an arch bridge failure. A research project to

develop load design procedures for bridges under debris flow impacts has

been submitted by the first author of this book. Explicit examples of arch

bridge damages or failures either due to rock falls or due to avalanches are

not known, however in general the failure of bridges due to such loads is

well-known (Proske 2009).

Besides natural accidental loads, technical load may also be applied to

arch bridges such as car, railway or ship impacts, or bombardment. The

problem of ship impacts against arch bridges has been intensively dis-

cussed in Proske (2003). Further discussion of arch bridge failures can be

found in Ural et al. (2008).

Some further examples from the last few decades are also mentioned.

The first example is the partial failure of the Molins de Rei bridge close to

Barcelona, Spain on 7th February 1971 and on 1st January 1972 (Troyano

2003). Pictures of the structure after the failure are shown in Troyano

(2003). On 9th April 1978, 6 of the 15 arches of the Wilson Bridge in

Tours, France collapsed (Troyano 2003, Rombock 1994). A further exam-

ple was the failure of the Westminster Bridge in Humberside Country 1983

(Tingle and Heelbeck 1995).

Besides structural damages, the building material itself can also be dam-

aged. Masonry, as already mentioned, is a multi component material. The

damages can therefore effect single elements alone, such as the mortar or

the stone, or they can effect the masonry. Translation of terms related to

masonry can be found on the ICOMOS (2008) web page or in Bau.de

(2008).

Mortar, the joint material of masonry, usually exhibits a much lower lifetime

and strength than the natural stones or masonry bricks. Natural stones can

reach hundreds or thousands of years of lifetime still keeping their strength

and showing only minor weathering effects.

However, the low weathering resistance of some mortar types can also

influence the stone material. The breakout of the mortar material enables

the penetration of humidity and acceleration of a further weathering of the

remaining mortar and stone or brick material. Figure 6-3 shows the principal

consequences of wrong pointing application and mortar weathering.

6.2 Damages on Historical Arch Bridges 223

1995)

tour scaling (Fig. 6-4). Based on the penetration depth, such damages are

classified into the following areas:

• Chipping of stone material either in convex or in concave shape

• Flaking or exfoliation in thin layers

• Spalling or detachment of crusts with stone layers of more than 10 mm

The cause of such damages is manifold. Classical weathering, loss of

binding agent, cracks caused by frost, bacteria as nitrificants, or salt attack.

These issues are widely discussed in literature such as Sauder and Wiesen

(1993), Beeger (1992), Poschlod (1990), Weiss (1992), or Bläuer (1992).

A very interesting example of spalling was presented by Mann. He re-

ported on spalling of masonry in a tunnel. The spalling was caused by the

smoke gas of the steam locomotive, which caused a chemical reaction of

the mortar towards gypsum.

A further example on weathered masonry surfaces can be found in Gar-

recht (1997).

224 6 Damages and Repair

many cases by their water solubility. Besides the solubility, salts can also

damage by hygroscopic water absorption. Here, salts feature a blasting ef-

fect. This blasting effect is caused by an increase in volume during the

changing of moist and dry crystalline phases of the salt. If the pore system

is already saturated, then the crystallization pressure can damage the struc-

tural material. Besides the crystallization pressure, hydration pressure can

also be observed. Water is then chemically attached to the salt in certain

temperature regions. Also, this process is characterized by a volume in-

crease. Detailed information about certain crystallization and hydration

pressures subject to saturation grade can be found in Weber (1993). A

short summary of damaging salts is given in Table 6-4.

6.2 Damages on Historical Arch Bridges 225

Weber (1993)

Class of chemical Name

compound

Sulphate compounds MgSO4 • 7 H2O Acrid salt

CaSO4 • 2 H2O Gypsum, calcium sulphate

Na2SO4 • 10 H2O Sodium sulphate

Nitrate compounds Mg(NO3)2 • 6 H2O Magnesium nitrate

Ca(NO3)2 • 4 H2O Calcium nitrate

5 Ca(NO3)2 • 4 NH3NO3 •

10 H2O

Chloride compounds CaCl2 • 6 H2O Calcium chloride

NaCl Common salt, sodium chloride

Carbonate compounds Na2CO3 • 10 H2O Sodium carbonate

K2CO3 Potash, Calcium carbonate

the damage by the salts is greater. However, the salts require humidity and

water as a transport medium and therefore mixed damages are common

(Weber 1993).

Damages related to humidity and salt are the following:

• Frost damage

• Spalling caused by hydraulic swelling and shrinkage

• Crystallization damage by salts

• Hydration damage by salts

• Frost-thaw damage

• Binding material reaction caused by acid exhausts

• Damage caused by microorganisms

Because damage by salts can be simply avoided by water penetration

exception, hydrophobicity of stones not only prevents water damage but

also salt damage. Hydrophobicity decreases the capillary suction capability

of materials. Most construction materials such as masonry or concrete suck

water on the surface. The wetting angle of contact is zero. Hydrophobicity

increases the wetting angle up to 90° or 180°. Since the capillary suction is

proportional to the cosine of the wetting angle, this yields to a cancellation

of capillary suction. However, this does not mean that the material is

sealed. If water with pressure is applied, this water can penetrate the material.

226 6 Damages and Repair

sonry material subject to different chemical reactions. Usually, moisture

and humidity are common requirements for such chemical reactions.

With the assimilation of moisture from the environment, usually other

chemicals are assimilated such as sulphur dioxide, nitric oxide, or carbon

dioxide. These elements then form acids and bases that solve the binding

material of the stones and mortar. The loss of binding material yields to

spalling, contour scaling, chipping, and flaking (Bartuschka 1995).

structures by biological processes. Such biological processes can be

microorganisms, moss growth, or the growth of plants and trees. Such

organisms either cause some chemical reactions or are able to introduce

stresses and forces inside the structural elements.

Mattheck et al. (1993) have measured the compression and tensile

strength of tree roots and found maximum compression stresses of up to

0.7 MPa, and tensile stresses of up to 50 MPa in longitudinal direction of

the tree root. Müller (2005) and Bauriegel (2004) investigate the strength

of trees under certain types of loadings on normal land. Although such

tests do not reflect in detail the conditions discussed here, they give a good

impression about the load transfer into roots. Garston (1985) has investi-

gated the influence of old trees to houses. Mattheck et al. (1993) give a

good example about the load-bearing capabilities of tree roots: in 1993 in a

northern German city, a tree root lifted up a gas pipe. This caused a gas

explosion.

tion material. For example, the coefficient of thermal expansion can yield

to different strains, causing different compression or tensile stresses inside

the material. If the stresses exceed the strength, then the material will crush

or crack. Such cracks can accelerate the weathering in combination with

water and salt penetration. However, physically caused damages are gen-

erally of minor importance for historical arch bridges (Bartuschka 1995).

6.2 Damages on Historical Arch Bridges 227

6.2.9 Deformations

to perform. However, if the magnitude of the deformations is too high, de-

formations can be considered as damage. The line between common defor-

mations and damage is difficult to find under practical conditions. For ex-

ample, arch bridges have already shown significant deformations after the

destruction of the falsework. Early bridges in the 18th century showed ver-

tical deformations in the crown of more than a 1 cm per m span. The span-

drel walls were often completed half a year after the finishing of the arch.

At that time, nearly 1/3 of the creep deformation of the arch had already

occurred. Bridges constructed in the second part of the 19th century did

not reach such high vertical deformations (usually between 0.1 and 0.4

mm/m). This can probably be related to an increased mortar quality

(Brencich and Colla 2002).

A good example to illustrate deformations in arch bridges is the Syratal

Bridge in Plauen, Germany. The deformation at the crown reached 55.57 cm

in 1995. The bridge has a span of 90 m. It has been assumed that the crown

deformation will reach 56.90 cm by 2070. This represents a ratio of nearly

0.57/90 = 6/1,000. The temporal development of the deformation is shown

in Fig. 6-5. The cause of the high deformation is manifold. Figure 6-6 tries

to relate certain causes to certain deformation values.

Perronet already knew about the great deformation of arch bridges. At

the Neuilli Bridge in Paris (1782–1783), he measured a deformation of 0.7

cm per m after the destruction of the falsework. He assumed that this value

represented about 60% of the overall assumed deformation. In the next 12

months, a further 30% of the overall deformation was observed and the fi-

nal deformation was found after five years (Brencich and Colla 2002).

Weber (1999) gives a deformation of 66 mm for the Lavour Bridge after

removal of the falsework. Harvey (2006) mentions deformation at the

springing of 0.1 mm under traffic load. He indicates that stones can come

off in the section of maximum traffic load.

According to Brencich and Colla (2002), high deformations of the arch

can cause cracks in hidden vaults and in the spandrel walls. Such cracks

yield to a separation, and therefore the ultimate load-bearing capacity of

the bridge may be changed due to the changed interaction of the single

structural elements.

228 6 Damages and Repair

Fig. 6-5. Crown deformation of the Syratal Bridge, Plauen over time (Span 90 m)

(Bartuschka 1995)

Fig. 6-6. Contribution of different causes to the crown deformation of the Syratal

Bridge, Plauen (Bartuschka 1995)

6.2 Damages on Historical Arch Bridges 229

6.2.10 Cracks

Cracking is caused by the exceedance of the tensile strength of a material.

Cracks are common phenomena in brittle, low tensile strength materials.

Such materials are, for example, glass, natural stones or concrete. Single

cracks are not necessarily a damage as seen with reinforced concrete. Here,

the concrete has to crack to permit the steel reinforcement contribution in

the load bearing. In such cases, the cracking is considered during the de-

sign process and only the crack size has to be limited.

Certain types of cracks in masonry are shown in Fig. 6-7. Such crack

patterns are strongly related to the failure surfaces as discussed in Chapter 3.

However, looking more towards masonry arch bridges, further classifica-

tion of cracks seems to be useful.

Cracks in the arch or vault can be an indication of overload on the struc-

ture. Since the geometrical location of the crack permits further interpre-

tation, certain types of arch bridge cracks are introduced in Fig. 6-8 and

classified. The comments of the cracks are mainly taken from Bienert

(1976), Bartuschka (1995), and the UIC-Codex (1995).

230 6 Damages and Repair

Fig. 6-7. Typical crack patterns in masonry walls according to Al Bosta (1999),

Jäger (2006), and Walthelm (1990, 1991)

6.2 Damages on Historical Arch Bridges 231

Fig. 6-8. Typical crack types in masonry arch bridges according to Bienert (1976)

Longitudinal cracks run parallel with the span of the bridge. They can ei-

ther run over the entire span of the bridge or cover parts of the bridge.

Since cracks always indicate tensile forces rectangular to the crack direc-

tion, longitudinal cracks indicate tensile forces in transversal direction of

the bridge. Such tensile forces can be caused by one-sided settlement of

the bridge, which is the main cause, while transversal bending can be

caused by one-sided traffic load on wide arches or vaults with several

lanes or high shrinkage stresses on wide arches. Longitudinal cracks often

indicate damage to the sealing. In general, longitudinal cracks do not indi-

cate an immediate threat to the ultimate load-bearing capacity of the

bridge. Longitudinal cracks are, for example, mentioned in Boothby et al.

(2004).

232 6 Damages and Repair

Front circle cracks or front ring cracks are a special type of longitudinal

cracks in arch bridges. Front circle cracks are located directly behind the

spandrel wall in the arch or vault. Sometimes they reach down to the piers

or abutment. Usually, the causes are different: stiffnesses of the spandrel

wall and the backfill, high traffic loads which cause movements of the

spandrel wall, or moisture penetration caused by damaged sealing. In con-

trast to the good-natured longitudinal cracks, front circle cracks may cause

a distinct change of the load-bearing behaviour of the arch bridges because

the spandrel wall separates from the arch. Usually, the spandrel walls con-

tribute strongly to the load bearing of the arch and therefore the loss of this

contribution has strong effects on load-bearing behaviour. The question

then arises whether the spandrel wall has been considered in the static

computation of the arch bridge (Bartuschka 1995).

Extrados joint cracks or arch back cracks are longitudinal cracks located at

the back side of the arch. They also yield to a separation of the spandrel

wall from the arch. However, the consequences are lower compared to the

front circle cracks since they indicate an original weak interlook between

the spandrel wall and the arch. Under such conditions, the load capacity of

the spandrel wall cannot be considered anyway since the interaction be-

tween the spandrel wall and the arch was not realized during design and

construction. Causes of such cracks can be differences of stiffness between

arch and spandrel wall, differences of stiffness between arch and the back-

fill, shrinkage deformations, or high horizontal loading inside the backfill

caused by traffic load or frost (Bartuschka 1995).

Transversal cracks run rectangular to the span of the arch. They occur

mainly at the springing, at the quarter point of the arch or at the crown, and

indicate the development of hinges in the arch (Fig. 6-9). Therefore, they

are a serious sign of overloading of the arch. Further causes besides the

overloading of the arch can be settlement, introduction of high single loads

with low coverage, improper shape of the arch, or high shear stresses in the

horizontal working joints of the backfill.

6.2 Damages on Historical Arch Bridges 233

Fig. 6-9. Transversal cracks caused by horizontal movements of the support ac-

cording to Como (1998)

weakness (Bauriegel 2004)

of the support can be found in Ochsendorf (2002) and Ochsendorf et al.

(2004). A more general introduction to the consequences of settlement for

historical structures was given by Bauriegel (2004). As an example, dam-

ages to vault constructions are shown in Fig. 6-10.

234 6 Damages and Repair

Diagonal cracks appear rather seldom in arches and vaults. If such cracks

are found, they have to be inspected. Causes for the development of the

cracks can be local weakness of the masonry or unequal load distribution.

The unintended load bearing of the spandrel wall at the springing can

cause diagonal cracks in the spandrel wall (Bartuschka 1995).

Sometimes displaced stones can be found in the arch. The reason why the

stones have moved has to be investigated very carefully. In most cases,

such moved stones occur on arch bridges with low coverage, high single

loads, and low bound between the stones. Such stones can endanger traffic

and humans under the bridges (Bartuschka 1995, Harvey 2006).

Spandrel walls and parapets can show some special types of damage.

Melbourne (1991) has classified the damage as shown in Fig. 6-11. An ex-

ample of sliding is shown in Fig. 6-12. Several examples of the complete

failure of spandrel walls by overturning after earthquakes can be found in

Rota (2004).

Como (1998) describes crack patterns in spandrel walls of arch bridges

caused by settlement of the middle pier (Fig. 6-13). Fauchoux and Abdunur

(1998) have indeed found such crack patterns at a bridge with pier settle-

ment. They have repaired the bridge by removal of the backfill, needling

of the arch masonry, and inserting a new concrete backfill.

Fig. 6-11. Different damage types on spandrel walls and parapets according to

Melbourne (1991)

6.3 Repair and Strengthening 235

Fig. 6-13. Crack pattern on spandrel walls caused by vertical settlement of the

middle pier according to Como (1998)

6.3.1 Introduction

Repair is a part of the maintenance of structures. According to DIN 31 051

(2003), all actions to conserve and restore the normal conditions and to

236 6 Damages and Repair

investigate and assess the actual conditions are integral parts of what is

called “maintenance.” Therefore, maintenance includes all types of inspec-

tion, servicing, and repair. For example, it includes damage and failure in-

vestigation, undertaking mitigation measures, repairing and mending, re-

placement and assembly, testing, and clearance. Inspection itself includes

all means to investigate and assess the actual conditions of a system. That

includes testing, measuring, assessment, and documentation. Servicing in-

cludes all actions to keep a system in normal conditions. For example, test-

ing, adjusting, exchanging, supplementing, preserving, and cleaning are

parts of servicing. It seems to be meaningful to link inspections and servic-

ing. Therefore, often the costs for both are given together (Curbach et al.

2003b).

Table 6-5 gives some German rules for investigation periods of bridges.

In contrast, Table 6-6 gives some indications for the maintenance planning

for arch bridges.

DIN 1076 RiL 804 Time distance

Ongoing observation Observation Permanent, half-yearly

Observation ------------------- Yearly

Simple inspection Inspection 3 years

Main inspection Expertise 6 years

Inspection caused by event Special inspection

Table 6-6. Return period of different maintenance actions on arch bridges accord-

ing to Steele et al. (2006)

Maintenance activity Every (years)

Vegetation removal 5

Coping stone replacement/realignment 10

Brickwork maintenance – repoint/renewal 15

Parapet repairs/replacement 15

Invert clearance 20

Cutwaters replaced 40

First refurbishment scheme 120

Second refurbishment scheme 200

introduce a maintenance factor:

R0,R = M ⋅ R0 . (6-1)

6.3 Repair and Strengthening 237

Table 6-7. Statistical properties of factor M according to Melchers and Faber (2001)

Quality of maintenance Mean value Standard deviation

Low 0.90 0.10

Good 0.95 0.05

Excellent 1.00 0.02

a structure. In contrast, a low quality of maintenance restores the load-

bearing capacity only to 90%. However, this model is rather simple. In

many cases, some maintenance actions restore the capacity, whereas other

actions decrease the capacity by structural work. In many cases the load-

bearing capacity of historical arch bridges also has to be increased.

Orbán (2004) has published an investigation about common repair

works on masonry railway arch bridges (Table 6-8). Not only does it list

the different methods but it also indicates the frequency of the certain

methods. Rombock (1994) mentions the following maintenance works for

natural stone material:

• Cleaning and hydrophobicity of natural stones

• Treatment of salt damages on natural stone masonry

• Conservation of natural stones

• Chemical stone cleaning

• Drying of masonry

• Mechanical cleaning

• Strengthening of natural stones

• Supplement of natural stones

The list of refurbishment techniques suggested by Page (1996) is listed in

Table 6-9.

Table 6-8. Repair techniques for masonry railway arch bridges according to

Orbán (2004)

Repair technique Percent of railway organi-

zations with experience in

the repair technique (%)

Maintenance of sealing:

Drainage pipes re-positioned and put through the arch 58

New backfill and roadway slab with sealing 42

Sealing without bond on the arch 33

Grouting of cement and micro-cement into the arch or 25

vault

Grouting of gel through the arch 17

238 6 Damages and Repair

zations with experience in

the repair technique (%)

Increase of load-bearing capacity –

Injection into the arch or vault 83

Shotcrete at the introdos of the arch 58

New backfill and roadway slab on the vault 42

Nailing of cracks with grouting of the nails 33

Support of the vault by steel arches 25

Increase of load-bearing capacity of the abutments, –

foundations and piers

Piles through the abutment 67

Nailing and grouting 50

Protection against erosion (sheet pile, concrete cover, 42

stone plasters around the pier)

Addition of reinforced concrete elements 33

Injection into the ground 33

Introduction of load-bearing capacity into the width of

the arch

Tie road and anchor plates 67

Connection between spandrel wall and arch 17

Reinforced concrete slab on the arch 25

Shotcrete at the intrados and connection of the spandrel 8

walls on the concrete by tensile elements

Table 6-9. Maintenance on masonry arch bridges according to Page (1996) and

COST 345 (2006)

Fault Measure

Deteriorated pointing Repoint

Deteriorated arch ring Repair masonry

Install saddle

Apply sprayed concrete to intrados

Install pre-fabricated liner

Grout arch ring

Apply proprietary repair technique

Arch ring inadequate to carry in- Install saddle

service loads Apply sprayed concrete to intrados

Install pre-fabricated liner

Replace fill with concrete

Install steel beam-relieving arches

Install relieving slab

Apply proprietary repair technique

Internal deterioration of mortar, Grout arch ring

which could lead to ring separation, Stitch (using tie bars spanning across a

for example crack)

6.3 Repair and Strengthening 239

Fault Measure

Foundation movement Install mini-piles or underpin

Grout piers and abutments

Outward movement of spandrel walls Install tie bars

Install spreader beams

Replace fill with concrete

Demolish walls and rebuild

Grout fill

Separation of arch ring beneath span- Stitch together

drel wall from remainder of arch ring

Weak fill Replace fill with concrete

Grout fill

Reinforce fill

Water leakage through arch ring Make road surfacing water resistant

Install waterproofing

Waterproof extrados and improve drainage

Scour, or damage to scour protection Install protection measures

works

Repair or enhance protection system, for

example by placing riprap or concrete

around substructures at risk

restore the load-bearing capacity of the arch:

• Replacement of the backfill

• Application of shotcrete shell

• Construction of a bridge inside the bridge (see for example Notkus and

Dulinskas 2002 and Stritzke 2007 and Fig. 6-14)

• Construction of a railway reinforced concrete slab

Fig. 6-14. Example of a bridge in bridge construction taken from Stritzke (2007)

tion phase. Both sequences are shown in Figs. 6-15 and 6-16.

Furthermore, a summary of certain reconstruction techniques can be

found in Hamid et al. (1994) or at Mathur et al. (2006).

240 6 Damages and Repair

ment cases on historical arch bridges built with natural stones. Standfuß

and Thomass (1987) state that the view of historical arch bridges should

be changed by only minor refurbishment. For example, shotcrete and con-

crete layers such as shown in Knoblauch et al. (2008) should only be applied

6.3 Repair and Strengthening 241

that under nearly all conditions, the so-called replacement of lay bricks can

be applied to re-establish the original conditions. Replacement of lay

bricks considers the replacement of damaged masonry parts by construct-

ing a falsework that supports the arch and lay bricks in the damaged re-

gions after replacing the damaged parts. For this technology, comparable

natural stones should be applied. Furthermore, sealing of the historical

arch bridges is a must according to Standfuß and Thomass (1987).

Mabon (2002) considers the construction of a concrete backfill and the

reinforced concrete railway slab as the most popular techniques for the

strengthening of historical natural stone arch bridges. Mabon (2002) fur-

thermore mentions the application of reinforcement in cut slots and of a

shotcrete layer.

Vockrodt (2005) and Vockrodt et al. (2003) also mention some restora-

tion examples in detail. Witzany et al. (2008) describe experimental stud-

ies in strengthening techniques.

Certain types of damage mechanisms were given in this chapter. One major

type was chemical attacks, such as salt penetration. Measures against salt

can be classified into chemical and physical ones. The chemical measures

mainly use the idea to transform aggressive salts into non-aggressive salts.

However, since different types of salts can be found inside the stone under

realistic conditions, it seems to be improbable to transform all aggressive

salts into harmless ones.

The physical salt decontamination uses intermediate plasters. Unfortu-

nately, such a technology seems to be impracticable for bridges. Addition-

ally, electro-physical technologies based on electro-osmosis are known.

Under all conditions, a cleaning of the natural stones should be carried

out. Table 6-10 lists and relates certain types of cleaning technologies to

certain types of stones. Local damages of stones can be repaired by reim-

2

bursement. For example, for stone damage zones smaller than 200 cm ,

restoration mortar can be used. If the damage zones are greater, then for

the reimbursement natural stone parts should be used. Bartuschka (1995)

gives the following working steps:

1. Pick out at least 2 cm deep and dovetail shaped (Fig. 6-17).

2. If the damage zones are greater, stainless steel reinforcement should be

built in. The reinforcement should not run over joints.

3. The surface of the reimbursement should be prepared by a stone cutter.

242 6 Damages and Repair

Table 6-10. Cleaning technologies for certain natural stones according to Bartuschka

(1995)

Cold water cleaning without

Cleaning compound

Steam jet cleaning

Sand blasting

Sandstone Beebly pressure

{ + + – {

Limy { { + – +

Clayey { + + – +

Lime stone Absorbent, soft { { + – {

coarsely porous + { + { {

coarsely porous, buffed + { + – {

Dense + { + { {

Dense, buffed + { + – +

Granite, Diorite + { + { +

Syenite, Labradorite (buffed) + { + – +

Tuff { { + – +

Marble Not buffed + { + – +

Buffed + { + – +

crystalline schist (not buffed) { { + { {

Phyllite, Serpentinite (buffed) { { + – +

Brick Not glazed { { + – +

Glazed { { + – +

Fig. 6-17. Examples of stone picking out according to Bartuschka (1995)

6.3 Repair and Strengthening 243

Many works have been published about the renovation of natural stones.

Further references of Ruffert (1981), Nodoushani (1997, 1998), Wihr

(1980), Reul (1994), Bienert (1976), and Sauder and Wiesen (1993) are

given.

The renovation of the stone material should always further include a res-

toration of the joint material since both act together. Furthermore, the joint

mortar often damages the stone material.

Mortar refurbishment mainly includes the removal of loose mortar and the

building of new joint material. This can be done either by hand or by

pressing by force. Several different techniques are shown in Bartuschka

(1995) and Jäger (2006). For example, the steps for the dry injection

method are shown in Fig. 6-18.

d Open and weakened joint

e Backfill

f Hammer

g Cleaned weak joint mortar

d Open and weakened joint

e Backfill

f Jet nozzle

g Cleaned weak joint mortar and

jet material

244 6 Damages and Repair

d Open and weakened joint

e Backfill

f Jet nozzle

g Trass lime mortar for joint tuck-

pointing

strengthening technology (Fig. 6-19). Miri and Hughes (2004) have carried

out tests on a scale of 1:12 which prove the increase impressively. The ap-

plication of a concrete slab increases the load-bearing capacity by a factor

between 3.2 and 3.7, depending on the ratio rise to span. In general, the

tests by Miri and Hughes (2004) have shown the load-deflection curve in

Fig. 6-20. It should be mentioned here that a former German railway rec-

ommendation gave an increase in the load-bearing capacity without any

detailed computation of a factor 1.2. The increase can be easily explained

by the change of the loading mechanism in arches. See the change of beam

models in Chapter 3 from simple arch models toward the model of Gocht

(1978).

6.3 Repair and Strengthening 245

Fig. 6-20. Qualitative deformation of an arch with and without reinforced concrete

slab according to Miri and Hughes (2004)

246 6 Damages and Repair

Injections and grouting can be applied to masonry bridges for different

reasons. For example, the behaviour of the masonry can be homogenized,

sealed, and hollow sections can be closed.

If sealing is reached by injection, then the injecting material does have to

fulfill the requirements for not only the injecting process but also sealing

properties. From a technological point of view, the injection material

should have a high penetration capability. Therefore, usually a low mo-

lecular material is used. Additionally, mainly true fluid solutions are com-

mon. Emulsions and suspensions are only rarely used. Since the 1980s, be-

sides cement slurry and cement suspensions, further injection material has

become widespread. Such materials are

• Alkali silicate dissolutions

• Alkali methyl silicone dissolutions

• Combination of alkali silicate dissolution and alkali methyl silicone

dissolutions

• Alkali propyl silicone dissolutions

• Silane and low-molecular oligomer siloxane in organic dissolvent

• Water soluble silicone microemulsions concentrate

• Bitumen solution and melting mass

• Bitumen emulsion

• Organic resin in organic dissolvent

• Alkanes

However, injections and grouting include a great uncertainty subject to

the effectiveness and any long-term effects. Therefore, before the applica-

tion is launched, the technology has to be evaluated regarding quality as-

surance. To provide that, usually test injections are carried out and can be

used for the assessment of the injection value. Several months after the test

injection, a destructive and nondestructive test should be applied to inves-

tigate the quality and effectiveness of the injection and grouting. That in-

jection and grouting can be successfully applied, for example, has been

shown by Schueremans et al. (2003).

6.3.2.5 Reinforcement

Already grouting and injection are technologies for improving the me-

chanical material properties of the construction material – here, masonry.

To this class also belongs the technique of reinforcement. However, the

volume or area ratio is rather low. The technique concentrates much more

on a well-selected location of the reinforcement inside the masonry. This

requires an understanding of the load path flow inside the masonry. Examples

6.3 Repair and Strengthening 247

or for the strengthening of the arch can be found in Figs. 6-21 and 6-22.

In general, the reinforcement can be distinguished either according to

the material used in metallic and nonmetallic reinforcements, or according

to the forces that should be partially covered by the reinforcement such as

shear force reinforcement or bending moment reinforcement.

A special example of the application of reinforcement is given in the

next section.

(1997) and COST 345 (2004)

248 6 Damages and Repair

Fig. 6-22. Reinforcement concept for arches according to Woodward (1997) and

COST 345 (2004)

The Archtec technique has been applied more than 130 times in Great

Britain, the United States, and Australia since 1998 (Brookes and Mullet

2004). As a specific example, the Wisconsin Avenue Bridge in the United

States is mentioned (Darden and Scott 2006)

The general idea of the technology is the strengthening of arch segments

by additional reinforcement elements. Since the most frequent failure of

arches is the development of mechanisms, including hinges at the quarter

points, the application is applied there to increase the bending moment ca-

pacity in the cross sections. However, in contrast to other strengthening

measures, such as replacement of the backfill by concrete, here no massive

construction works have to be undertaken at the bridge. Rather, only

drillings are carried out and then the reinforcement such as threaded rods

is applied into the drilling holes. Figure 6-23 shows, as an example, the lo-

cation of reinforcement elements with a length of 2.5 m. The effective-

ness of this concept has been proven in tests where the reinforcement ele-

ments as well as the arch were equipped with strain-measuring devices. Of

course, the strengthening can only be active for traffic load. The dead load

6.3 Repair and Strengthening 249

Fig. 6-23. Profile of an arch bridge with the position of the reinforcement ele-

ments

still has to be taken entirely by the original arch (Brookes and Mullet 2004,

Mabon 2002, Owen et al. 2005 and Tilly and Brookes 2005).

Oliveira and Lourenço (2004) introduced the installation of transversal re-

inforcement elements into arches (Fig. 6-24) or above the extrados. The

elements are designed to overtake tensile forces in transversal direction.

The reinforcement elements are profile bars. A comparable solution has

been introduced by Falconer (1999). He anchored the reinforcement above

the arch stones. In Germany, model drawings and recommendations for

the anchoring of reinforcement elements in the spandrel walls exist

(BMVBW 1993). A further example can be found in Welch (1995).

Fig. 6-24. Transversal reinforcement elements in an arch taken from Oliveira and

Lourenço (2004)

250 6 Damages and Repair

Besides the application of classical steel reinforcement, nonmetallic rein-

forcement elements can also be applied. The major advantage is the exclu-

sion of corrosion. Such techniques are applied, not only to natural stone

bridges, but also to classical steel reinforced concrete elements. The first

author has used textiles for reinforcement for concrete slabs (Proske 1997).

Other materials have also been applied to arch bridges.

For example, Modena et al. (2004) report about the installation of car-

bon fibre-reinforced polymer (CFRP) elements transversally and longitu-

dinally into arches. Melbourne and Tomor (2004) and Hodgson (2003)

have also used CFRP elements for arch strengthening.

Bergmeister (2003) introduces the strengthening of a concrete arch

bridge in transversal direction by CFRP elements.

Fiber-reinforced polymers (FRP) have been applied for the strengthen-

ing of historical arch bridges by De Lorenzis and Nanni (2004), Creazza

and Saetta (2001), Valluzzi and Modena (2001), Borri et al. (2002) – Fig.

6-25, Bati and Rovero (2008), Ricamato (2007), Drosopoulos et al. (2007),

and Panizza et al. (2008).

Fig. 6-25. Strengthening example of arch bridge using FRP by Borri et al. (2002)

Pre stressed and nailed masonry applied to historical arch bridges will not

be discussed here in detail. For details, see Ganz (1990), Ullrich (1989),

and Wenzel (1997).

6.3.2.10 Shotcrete

The application of shotcrete with historical masonry arch bridges is not

recommended due to conservation criteria for monuments and historical

structures. However, under some conditions, it cannot be avoided, or it can

6.3 Repair and Strengthening 251

ings and recommendations exist about the application of shotcrete at the

intrados (BMVBW 1993).

In the section on damages, damages on spandrel walls and parapets was

also mentioned. Here, restoration techniques have been developed with ei-

ther using some hidden application of strengthening material, such as rein-

forced concrete, or using a decoupling of the backfill and the spandrel

walls. Examples are shown in Figs. 6-26 and 6-27. Further details can be

found in the COST 345 report (2006).

Fig. 6-26. Reinforced parapet according to Welch (1995) and COST 345 (2006)

252 6 Damages and Repair

Fig. 6-27. Decoupling of the spandrel wall according to Welch (1995) and COST

345 (2006)

An insufficient width of historical bridges is a major problem for the ad-

aptation of such structures to modern traffic requirements. Therefore,

widening of the bridges is common (Fig. 6-28). For example, in the last

few years in Spain, one fifth of all highway bridges have been widened

(Angeles-Yáñez and Alonso 1996). According to Harrison (2004), the

major cause for destruction of historical bridges in Great Britain was insuf-

ficient width of the road track.

An example of the successful widening of a historical arch bridge can be

found in Troyano (2003). Here, the widening of the Pont Vieux de Albin

over the river Tarn and Burgo in Spain was carried out using a low-pitched

arch. Another example was given by Parikh and Patwardhan (1999). The

widening of the Marienbrigde in Dresden was described by Koettnitz and

Schwenke (1998). See also Scheidler (1992), Sobrino (2007), Boronczyk-

Plaska and Radomski (2008), and Vockrodt et al. (2003).

A very special example is the Chemnitz Viaduct in Germany, where the

span of the arch was too low for the highway traffic under the bridge.

Therefore, one pier was supported by a bridge under the historical arch

bridge (Fig. 6-29). Also, the New Saale Bridge South Jena on German

highway A4 is mentioned here (Martin and Becker 2005). Although the

bridge is a new reinforced bridge to provide sufficient overall width for the

highway, it uses the shape of the historical arch bridge in which

neighbourhood it is built (Fig. 6-30).

6.3 Repair and Strengthening 253

strategies applied to historical arch bridges. Therefore, the following sec-

tion summarizes some refurbishment examples for historical arch bridges

but does not intend to be a complete list. However, it should give an im-

pression about the problems faced under practical conditions.

Fig. 6-29. Chemnitz Viaduct after removal of the piers (Germany) according to

Reintjes (2002)

254 6 Damages and Repair

6.3.3 Examples

The historical bridge over the river Werra in the city of Münden in Lower

Saxony, Germany was found to show an insufficient load-bearing capacity

after a major bridge investigation. This assumption was based on erosion

found at the piers, heavy curvatures and shifts of the pier and arch ma-

sonry, efflorescence and strong weathering of the masonry joints, and fi-

nally strong corrosions of the steel anchors, iron clips, and the iron parapet

(Schwartz 1988).

Due to the historical importance of the bridge, a major maintenance ac-

tion was launched. The maintenance plan included, for example, the substi-

tution of the iron parapet by a massive parapet. Furthermore, the arch was

cleaned by high-pressure cold water. Damage to the stones was repaired by

substitution using natural stone material and small damage was repaired

using Mineros stone mass. This material includes stone flour, which is em-

bedded into a two-component synthetic (Schwartz 1988).

Another example is the refurbishment of the Taubern Bridge in Lauda.

The historical three-arch bridges were erected in 1512 with a span between

6 and 7 m. Due to the increasing traffic load, spandrel walls and wing

walls were moved and the masonry arch showed wide cracks. To provide

safety for the bridge, anchors and clips were built in. In the 1930s, a

steel jacket was installed at the extrados. However, the amount of dam-

age increased, and in the 1960s the weight restriction of the bridge was

intensified from 16 to 9 tonnes. Finally, the bridge was demolished and

6.3 Repair and Strengthening 255

concrete piles, the span and, very importantly, the width of the recon-

structed bridge were changed in relation to the original one. To conform at

least partially to conservation rules, parts of the original bridge were used

for the reconstruction. For example, the coverage of the bridge was carried

out using natural stone, sometimes even original parts. Also, for the bridge

platforms, the original bridge crucifixes were used (BMV 1988).

The third example is the Hoch Bridge Dingolfing. The bridge was origi-

nally constructed in 1612 as a five-span brick masonry bridge with single

spans between 5.40 and 6.35 m. The piers reach a width of 1.20 and 1.35

m. The bridge reaches an overall length of 54 m and a maximum height of

5.6 m. The brick material showed heavy damages due to long moisture

penetration.

At several locations, stones had separated from the masonry. Settlement

of the piers had yielded to cracks in the spandrel walls. The bridge was

maintained in 1750, 1850, and 1890. The latest refurbishment was carried

out in 1966. The refurbishment had contained a complete disassembly and

reconstruction of the spandrel walls, clearage of the backfill and refilling

with lean concrete, and installation of a sealing and drainage system. Fur-

thermore, special bricks were produced for the reconstruction, the remain-

ing elements were cleaned by sandblasting, and afterwards the joints were

filled again. For this filling, a special mortar was designed using pit lime

mortar with tuff additive. Additionally, the bridge-in-bridge system was

applied because the arch was separated from the reinforced concrete slab

by a reinforced concrete structure that carried the load directly from the

railway slab towards the piers. The arch crown was separated from the

concrete slab by a 5 cm strong polystyrene layer (BMV 1988).

The old Dreisam Bridge Eichstetten is a five-span basket arch bridge

with single spans between 4.70 and 5.00 m. The piers have a width of 1.35

m. The width of the bridge reaches 4.6 m. The arches are covered with

natural stone ashlar masonry, however the inner parts are only built with

quarries. The spandrel walls are also made of ashlar masonry. The parapet

again is built with quarries. In 1950, the bridge was refurbished by building

in a backfill with lean concrete and adding a reinforced concrete railway

slab. The spandrel walls were secured by tendons with steel anchors. Such

steel anchors were also used at the Kocherbridge Griesbach (BMV 1988).

The Wurm Bridge in Hessia, constructed in 1777, was maintained in

1979 only by adding a reinforced concrete slab and a new sealing to the

structure. This five-span bridge built of new red stone masonry, and with

spans between 3.0 and 4.5 m, is one of the few bridges in Germany not

blasted at the end of World War II (BMV 1988).

256 6 Damages and Repair

using new red sandstone. The bridge reached an overall length of 51.30 m

with four segmental arches spanning between 7.30 and 13.00 m. The piers

reached a width between 3.00 and 5.40 m. In 1852 and 1855, the bridge

was strongly maintained. In 1914, the width of the bridge was extended

from 5.0 to 12 m. This was done by constructing reinforced concrete

arches covered with natural stones (BMV 1988).

Many further examples can be found in the literature. As mentioned in

Chapter 1, the number of arch bridges still functioning in the infrastructure

is overwhelming. The efforts to keep such an essential piece are repre-

sented in many papers dealing with the strengthening of arch bridges.

Some examples are mentioned in Koettnitz and Schwenke (1998), Günther

et al. (1999), Vockrodt (2005), Patzschke (1996), and Zahn (1999). Recent

examples can be found in Fotheringham (2008), Asmar et al. (2008), Beben

and Manko (2008), and Siwowski and Sobala (2008).

Although the maintenance efforts for arch bridges are low if the lifetime is

considered, Weber (1999) suggests an improvement of arch bridges and

call these bridges second generation stone arch bridges. Such new arch

bridges feature the following:

• Abandonment of back and lining masonry above the extrados. This

would yield to an improved numerical description of the load-bearing

behaviour of the arch with lower construction costs

• Backfill material used should be cohesionless and coarse grained. Geo-

textiles or steel ribbons should be applied to limit the compression on

the spandrel walls. Sealing should be built in above the backfill

• Drainage is very important and therefore should be long lasting and

controllable

• Abandonment of spandrel walls depending on the conditions

• Application of new technologies to decrease falsework costs

• Usage of new developed natural stone stocks in Europe

• Application of automatic stone cutting techniques

Examples of new stone arch bridges are double-curved arch bridges in

China (1964). Another example is the Kimbolton Butts Bridge, a brick

stone arch bridge constructed in the 1990s in Great Britain.

Such new arch bridges have to fully comply with the safety require-

ments for modern structures.

References 257

References

und Schwinden. Baupraktische Beispiele. Werner Verlag 2. Auflage

Angeles-Yáñez M & Alonso AJ (1996) The actual state of the bridge management

system in the state national highway network of Spain. Recent Advances in

Bridge Engineering. Evaluation, management and repair. Proceedings of the

US-Europe Workshop on Bridge Engineering, organized by the Technical

University of Catalonia and the Iowa State University, JR Casas, FW Klaiber

& AR Marí (Eds), Barcelona, 15–17 July 1996, International Center for Nu-

merical Methods in Engineering CIMNE, pp 99–114

Asmar R, Buys JP, Deschamps D, Kassis B, Salame A, Slaha W, Bouchon E &

Vion P (2008) Expertise and repair of Maameltein arch bridge in Lebanon. In:

Forde: Structural Faults and Repair 2008. 12th International conference,

Edinburgh, 10–12th June 2008, Structural faults and repair 2008

Bartuschka P (1995) Instandsetzungsmaßnahmen an hochwertiger Bausubstanz

von Gewölbebrücken. Diplomarbeit. Technische Universität Dresden, Institut

für Tragwerke und Baustoffe

Bati SB & Rovero L (2008) Towards a methodology for estimating strength and

collapse mechanism in masonry arches strengthened with fibre reinforced

polymer applied on external surfaces. Materials and Structures, 41 (7), pp

1291–1306

Bau.de (2008) Bauen und Wohnen im Internet. www.bau.net/woerterbuch/

Bauriegel A (2004) Empfehlung Nr. 3 des Arbeitskreises Geotechnik historischer

Bauwerke und Naturdenkmäler der DGGT: Geotechnische

Schadensfeststellung und -behandlung an historischen Bauwerken. Bautech-

nik, 81 (9), pp 760–765

Beben D & Manko Z (2008) Modernization of old arch bridges by steel shell

structure made from corrugated plates. In: Forde: Structural faults and repair

2008. 12th International conference, 10–12th June 2008, Edinburgh

Beeger D (1992) Naturstein in Dresden, Taschenbuch, Staatliche Naturhistorische

Sammlungen Dresden, Museum für Mineralogie und Geologie

Bergmeister K (2003) Kohlenstofffasern im Konstruktiven Ingenieurbau. Verlag

Ernst und Sohn, Berlin

Bién J & Kamiński T (2004) Masonry arch bridges in Poland. Arch Bridges IV –

Advances in Assessment, Structural Design and Construction, P Roca & C

Molins (Eds), CIMNE, Barcelona, pp 183–191

Bień J & Kamiński T (2007) Damages to masonry arch bridges – proposal for

terminology unification. Proceedings of the Arch 07 – 5th International

Conference on Arch Bridges, PB Lourenço, DV Oliveira & A Portela

(Eds), 12–14 September 2007, Madeira, pp 341–348

Bienert G (1976) Brückenerhaltung. 2. Auflage Transpress Verlag

Bläuer C (1992) Mineralogische Schadenaufnahme und Interpretation der

Schadenursachen an der Nydeggbrücke in Bern. Straße und Verkehr.

Zeitschrift des Vereins Schweizerischer Straßenfachleute 2, pp 88–89

BMV (1988) Bundesminister für Verkehr: Steinbrücken in Deutschland. Beton-

Verlag Düsseldorf

258 6 Damages and Repair

Sofortinstandsetzungsmaßnahmen an Brücken und anderen

Ingenieurbauwerken der Bundesfernstraßen in den neuen Bundesländern;

Sammlung von Arbeitshilfen für die Planung und Vergabe, Verkehrsblatt-

Verlag

Boothby TE, Hulet HM & Stanton TR (2004) Inspection, assessment and monitor-

ing of Railroad arch bridges in Sothwestern Pennsylvania. Arch Bridges IV –

Advances in Assessment, Structural Design and Construction, P Roca & C

Molins (Eds), CIMNE, Barcelona, pp 144–151

Boronczyk-Plaska G & Radomski W (2008) Bridge widening – technical, eco-

nomical and aesthetical aspects. Bridge Maintenance, Safety, Management,

Health Monitoring and Informatics, H Koh & DM Frangopol (Eds), Taylor &

Francis Group, London, pp 2850–2857

Borri A, Corradi M & Vignoli A (2002) Seismic upgrading of masonry structures

with FRP. Universita’ Degli Studi di Perugia, Facoltà di Ingegneria

Bottesford Living History (2007)

www.bottesfordhistory.org/page_id__242_path__0p25p.aspx

Brencich A & Colla C (2002) The influence of construction technology on the

mechanics of masonry railway bridges, Railway Engineering 2002, 5th Inter-

national Conference, 3–4 July 2002, London

Brookes CL & Mullet PJ (2004) Service load testing, numerical simulations and

strengthening of masonry arch bridges. Arch Bridges IV – Advances in As-

sessment, Structural Design and Construction, P Roca and C Molins (Eds),

CIMNE, Barcelona, pp 489–498

Como M (1998) Minimum and maximum thrust state in statics of ancient masonry

bridges. Arch Bridges: History, Analysis, Assessment, Maintenance and Re-

pair. Proceedings of the Second International Arch Bridge Conference,

A. Sinopoli (Ed), Venice 6.–9. October 1998, Balkema, Rotterdam, pp 133–137

COST 345 (2006) European Commission Directorate General Transport and En-

ergy: Procedures Required for Assessing Highway Structures: Working Group

6 Report on remedial measures for highway structures.

http://cost345.zag.si/Reports/COST_345_WG6.pdf

Creazza G & Saetta AV (2001) Analysis of masonry structures reinforced by FRP.

Historical Constructions, PB Lourenço & P Roca (Eds), University of Minho,

Guimarães, pp 539–545

Curbach M, Bösche T, Scheerer S, Michler H, Proske D (2003a) Wiederaufbau

der Pöppelmannbrücke Grimma

Curbach M, Köster T., Proske D, Schmohl L, Taferner J & Ehmann J (2003b)

Parkhäuser, Betonkalender 2004, Teil II, Ernst & Sohn, pp 3–153

Darden C & Scott TJ (2006) Strengthening from Within. Federal Highway Ad-

ministration, Public Roads, http://www.tfhrc.gov/pubrds/05mar/07.htm

De Lorenzis L & Nanni A (2004) International Workshop on Preservation of His-

torical Structures with FRP Composites. Final Report. National Science

Foundation, USA, July 2004

DIN 1045-1 (2001) Tragwerke aus Beton, Stahlbeton und Spannbeton, Teil 1:

Bemessung und Konstruktion, Juli 2001

References 259

DIN 31051 (2003) Maßnahmen zur Verzögerung des Abbaus des vorhandenen

Abnutzungsvorrates

Drdácký MF & Slízková Z (2007) Flood and post-flood performance of historic

stone arch bridges. Proceedings of the Arch 07 – 5th International Conference

on Arch Bridges, Lourenço, Oliveira & Portela (Eds), 12-14 September 2007,

Madeira. pp 163-170

Drosopoulos GA, Stavroulakis GE & Massalas CV (2007) FRP reinforcement of

stone arch bridges: Unilateral contact models and limit analysis. Composites:

Part B 38, pp 144–151

Falconer RE (1999) Test on masonry arch bridges strengthened using stainless

steel reinforcement. Current and future trends in bridge design, construction

and maintenance in Bridge design construction and maintenance, PC Das, DM

Frangopol & AS Nowak (Eds), Thomas Telford, London, pp 415–423

Fauchoux G & Abdunur C (1998) Strengthening masonry arch bridges through

backfill replacement by concrete. Arch Bridges: History, Analysis, Assess-

ment, Maintenance and Repair. Proceedings of the Second International Arch

Bridge Conference, A Sinopoli (Ed), Venice 6.–9. October 1998, Balkema

Rotterdam, pp 417–422

Fotheringham N (2008) B9136 Ruthven bridge: Stone masonry refurbishment. In:

Forde: Structural Faults and Repair 2008. 12th International conference,

10–12th June 2008, Edinburgh

Ganz H-R (1990) Vorgespanntes Mauerwerk. Schweizer Ingenieur und Architekt,

Nr. 8. 22. Februar 1990, pp 177–182

Garrecht H (1997) Zum Verhalten bewitterter Mauerwerksoberflächen. In:

Universität Karlsruhe, Sonderforschungsbereich 315 “Erhalten historisch

bedeutsamer Bauwerke”, Jahrbuch 1994, pp 187–198

Garston (1985) The influence of trees on house foundations in clay soils, Building

Research Establishment Digest; Building Research Establishment, 298, pp

13–75

Gocht R (1978) Untersuchungen zum Tragverhalten rekonstruierter Eisenbahnge-

wölbebrücken, Dissertation Hochschule für Verkehrswesen “Friedrich List”

Dresden

Günther L, Schäfer J & Schwenke F (1999) Instandsetzung der Friedensbrücke

Bautzen. 9. Dresdner Brückenbausymposium, Technische Universität

Dresden, pp 123–150

HA (1997) Highway Agency. BA 16: The assessment of highway bridges. Design

Manual for Roads and Bridges, HMSO, London

Hamid AA, Mahmoud ADS & El Magd SA (1994) Strengthening and repair of

unreinforced masonry structures: State of the Art. Proceedings of the 10th IB2

MaC, Calgary, Canada, 5.–7.July 1994, pp 485–497

Harrison D (2004) The Bridges in Medieval England – Transport and Society

400–800. Clarendon Press, Oxford

Harvey B (2006) Some problems with arch bridge assessment and potential solu-

tions. Structural Engineer, 84 (3), 7th February 2006, pp 45–50

Hodgson JA (2003) The Use of CFRP Plates to strengthen Masonry Arch Bridges.

Extending the Life of Bridges, Concrete and Composites Buildings, Masonry

260 6 Damages and Repair

July 2003, London

ICOMOS (2008) International Scientific Committee for Stone (ICOMOS-ISCS)

http://lrmh-ext.fr/icomos/consult/consultation.php?lang=1

Jäger W (2006) Rissbilder und Putzabplatzungen bei Mauerwerksbauten –

Ursachen und Strategien zur Schadensbeseitigung, Lehrstuhl

Tragwerksplanung, Fakultät Architektur der TU Dresden

Knoblauch F-J, Laubach A, Sauer W, Schmidt M & Sprinke P (2008) Erneuerung

des Burtscheider Viadukts in Aachen. Beton- und Stahlbetonbau 103, Heft 3,

pp 195–203

Koettnitz R & Schwenke F (1998) Umbau und Instandsetzung der Marienbrücke

in Dresden. 8. Dresdner Brückenbausymposium, Technische Universität

Dresden, pp 139–156

Mabon L (2002) Assessment, strengthening and preservation of masonry struc-

tures for continued use in today’s infrastructures. IABSE symposium,

Melbourne

Martin R & Becker M (2005) Neubau Saalebrücke Süd bei Jena im Zuge der BAB

A 4 – Umsetzung der gestalterischen Vorgaben. 15. Dresdner Brückenbau-

symposium, 15. März 2005, TU Dresden, pp 115–136

Mathur M, Kumar SV & Singh K (2006) Assessment and retrofitting of arch

bridges. www.iricen.gov.in/projects/622/ARCH%20BRIDGE.pdf

Mattheck C, Tesari I & Bethge K (1993) Roots and Buildings. Structural Repair

and Maintenance of Historical Building III (STREMAH), CA Brebbia & RJB

Frewer. Computational Mechanics Publication, Glasgow, pp 751–760

Melbourne C & Tomor AK (2004) Fatigue Performance of Composite and Radial-Pin

Reinforcement on Multi-Ring Masonry Arches. Arch Bridges IV - Ad-vances in

Assessment, Structural Design and Construction, Roca and Molins (Eds),

CIMNE, Barcelona, pp 428–434

Melbourne C (1991) Conservation of masonry arch bridges. In: Proceedings of the

9th International Brick/Block Masonry Conference, 13.–16. October 1991,

Volume 1, Berlin, Germany, pp 1563–1570

Melchers RE & Faber MH (2001) Aspects of Safety in Design and Assessment of

Deteriorating Structures, Proceedings to the International IABSE Conference

on Safety, Risk and Reliability - Trends in Engineering, March 21-23, Malta

2001, pp 161–166

Meyer MG (2007)

www.spanien-bilder.com/aktuelles_aus_spanien_details3328.htm

Mildner K (1996) Tragfähigkeitsermittlung von Gewölbebrücken auf der

Grundlage von Bauwerksmessungen. 6th Dresdner Brückenbausymposium,

14th March 1996, Fakultät Bauingenieurwesen, Technische Universität

Dresden, pp 149–162

Miri M & Hughes TG (2004) Increased load capacity of arch bridges using slab

reinforced concrete. Arch Bridges IV – Advances in Assessment, Structural

Design and Construction, P Roca & C Molins (Eds), pp 469–478

Modena C, Valluzzi MR, da Porto F, Casarin F & Bettio C (2004) Structural up-

grading of a brick masonry arch bridge at the Lido (Venice). Arch Bridges IV –

References 261

Molins (Eds), pp 435–443

Müller P (2005) Biomechanische Beschreibung der Baumwurzel und ihre

Verankerung im Erdreich. Dissertation. Fakultät für Maschinenbau,

Universität Karlsruhe

Nodoushani M (1997) Instandsetzung von Natursteinbrücken. Beton-Verlag.

Düsseldorf

Nodoushani M (1998) Natursteingewölbebrücke (... über Fluss Sauer in

Grundsteinheim, Ortsausgang Richtung Igenhausen/Stadt Lichtenau...).

Bausanierung, Gütersloh 9, 4, pp 45–48

Notkus AJ & Dulinskas E (2002) Spatial non-linear analysis of complex structure

“bridge-in-bridge”. Journal of Civil Engineering and Management. Vilnius,

Technika, Volume VIII, Suppl 1, pp 29–34

Ochsendorf JA (2002): Collapse of Masonry Structures. University of Cambridge,

Dissertation, Cambridge

Ochsendorf JA, Hernando JI & Huerta S (2004) Collapse of Masonry Buttresses.

Journal of Architectural Engineering, ASCE, September 2004, pp 88–97

Oliveira DV & Lourenço PB (2004) Repair of Stone Masonry Arch Bridges. Arch

Bridges IV – Advances in Assessment, Structural Design and Construction,

Roca & Molins (Eds), CIMNE, Barcelona, pp 451–458

Orbán Z (2004) Assessment, reliability and maintenance of masonry arch railway

bridges in Europe. Arch Bridges IV – Advances in Assessment, Structural

Design and Construction, P Roca & C Molins (Eds), CIMNE, Barcelona, pp

152–161

Owen DR, Peric D, Petrinic N, Smokes C & James PJ (2005) Finite/Discrete Ele-

ment Models for Assessment and Repair of Masonry Structures.

www.cintec.com

Page J (1996) A guide to repair and strengthening of masonry arch highway

bridges. TRL Report 204, TRL Limited, Crowthorne

Panizza M, Garbin E, Valluzzi MR & Modena C (2008) FRP Strengthening –

shear mechanism of brick masonry vaults. In: Forde: Structural Faults and

Repair 2008. 12th International conference, 10–12th June 2008, Edinburgh

Parikh SK & Patwardhan YG (1999) Widening of an existing old masonry arch

bridge through RCC box girder deck. Current and Future Trends in Bridge

Design, Construction and Maintenance in Bridge Design Construction and

Maintenance, PC Das, DM Frangopol & AS Nowak (Eds), Thomas Telford,

London, pp 218–226

Patzschke F (1996) Der Wahrener Viadukt in Leipzig – Rekonstruktion einer

Gewölbebrücke der DB AG. 6. Dresdner Brückenbausymposium, 14.3.1996,

Fakultät Bauingenieurwesen, Technische Universität Dresden, pp 135–148

Poschlod K (1990) Das Wasser im Porenraum kristalliner Naturwerksteine und

sein Einfluss auf die Verwitterung. Dr. Friedrich Pfeil Verlag

Proske D (1997) Textilbewehrter Beton als Verstärkung von Platten, 34.

Forschungskolloquium des Deutschen Ausschuss für Stahlbeton, Technische

Universität Dresden, Institut für Tragwerke und Baustoffe, Dresden, pp 51–58

Proske D (2003) Ein Beitrag zur Risikobeurteilung alter Brücken unter

Schiffsanprall. Dissertation, Technische Universität Dresden

262 6 Damages and Repair

19th Dresdner Brückenbausymposium, 9–10 March 2009, Dresden

Reintjes K-H (2002) Die Unterfangung des Bahrmühlenviadukts Erfahrungen aus

der Bauabwicklung. 12. Dresdner Brückenbausymposium, 14. März 2002, TU

Dresden, pp 145–163

Reul H (1994) Handbuch Bautenschutz Bausanierung, Verlag Müller

Ricamato M (2007) Numerical and experimental analysis of masonry arches

strengthened with FRP materials. University of Cassino, Department of Engi-

neering, Graduate School in Civil Engineering, PhD thesis

Rombock U (1994) Restaurierung von Steinbrücken. Informationszentrum Raum

und Bau der Fraunhofer-Gesellschaft, 2., erweiterte Auflage, IRB-Verlag,

Stuttgart

Rota M (2004) Seismic Vulnerability of Masonry Arch Bridge Walls. European

School of Advance Studies in Reduction of Seismic Risk, Rose School, Dis-

sertation

Ruffert G (1981) Sanierung von Baudenkmälern, Beton-Verlag

Sauder M & Wiesen H (1993) Sanierung von Bauwerken aus Naturstein,

Taschenbuch, Landesinstitut für Bauwesen NRW

Scheidler J (1992) Umbau und Verbreiterung von Bogen- und Gewölbebrücken.

Würzburger Betonseminare der LGA Bayern am 11.–12. März 1992, Oberste

Baubehörde Bayerns

Schueremans L, Van Rickstall F, Ignoul S, Brosens K, Van Balen K & Van

Gemert D (2003) Continuous Assessment of Historic Structures – A State of

the Art of applied Research and Practice in Belgium, KULeuven, Department

of Civil Engineering

Schwartz J (1988) Sanierung der historischen Werrabrücke 1986–1988:

Schadensanalyse und Sanierungskonzept. Werrabrücken Münden –

Geschichte und Sanierung 1986–1988, pp 25–33

Siwowski T & Sobala D (2008) The revitalization of XIX century masonry arch

bridge. In: Forde: Structural Faults and Repair 2008. 12th International con-

ference, 10–12th June 2008, Edinburgh

Sobrino JA (2007) Extending the life of masonry and concrete arch bridges.

Structural Engineering International, 17 (4), pp 328–336

Standfuß F & Thomass S (1987) Grundsätze der baulichen Maßnahmen an alten

Steinbrücken. Bausubstanz, Nr. 3, p 28

Steele K, Cole G, Parke G, Clarke B & Harding J (2006): Application of life cycle

assessment technique in the investigation of brick arch highway bridges. Pro-

ceedings of the 2002 Conference for the Engineering Doctorate in Environ-

mental Technology,

http://www.cintec.co.uk/en/applications/Archtec/documents/Chapter05_4.htm

Strasser A (2006) Vorlesung Altern, Institut für Physiologie, Veterinärmedizinische

Universität Wien, Vienna

Stritzke J (2007) Bogen- und Stabbogenbrücken. In: Handbuch Brücken,

Mehlhorn (Ed), Springer Verlag, Berlin – Heidelberg

Switaiski B (2006) Ein systematischer Ansatz für eine Strategie zur Gewährleistung

der Sicherheit von Gebäuden. Proceedings of the 4th International

Probabilistic Symposium 2006, Medianpour (Eds), Gucma & Proske, BAM

Berlin.

References 263

www.cintec.com

Tingle B & Heelbeck M (1995) The Management of a new County’s Bridge Stock

Humberside 20 Years on. Arch Bridges, C Melbourne (Ed), Proceedings of

the 1st International Conference on Arch Bridges, Bolton, Thomas Telford,

London, pp 37–46

Troyano LF (2003) Bridge Engineering – A Global Perspective. Thomas Telford

UIC-codex (1995) Empfehlungen für die Bewertung des Tragvermögens

bestehender Gewölbebrücken aus Mauerwerk und Beton. Internationaler

Eisenbahnverband. 1. Ausgabe 1.7.1995

Ullrich M (1989) Statusbericht über Forschungsarbeiten zur ingenieurmäßigen

Sicherung von historischem Mauerwerk durch Verpressen, Vernadeln und

Vorspannen. Bautenschutz + Bausanierung 12, pp 19–27

Ural A, Oruc S, Dogangün A & Iskender Tuluk Ö (2008) Turkish historical arch

bridges and their deteriorations and failures. Engineering Failure Analysis 15,

pp 43–53

Valluzzi MR & Modena C (2001) Experimental analysis and modelling of ma-

sonry vaults strengthened by FRP. Historical Constructions, PB Lourenço &

P Roca (Eds), Guimarães, pp 627–635

Vockrodt HJ (2005) Instandsetzung historischer Bogenbrücken im Spannungsfeld

von Denkmalschutz und modernen infrastrukturellen Anforderungen. 15.

Dresdner Brückenbausymposium, 15. März 2005, Dresden, pp 221–241

Vockrodt HJ, Feistel D & Stubbe J (2003) Handbuch Instandsetzung von

Massivbrücken. Untersuchungsmethoden und Instandsetzungsverfahren.

Basel, Boston, Berlin: Birkhäuser Verlag

Walthelm U (1990) Rissbildungen und Bruchmechanismen in freistehenden

Wänden aus Beton und Mauerwerk. Bautechnik 67, Heft 1, pp 21–26

Walthelm U (1991) Theoretische Spannungszustände, Rissbildungen und

Traglasten von einachsig ausmittig gedrückten Rechteckquerschnitten.

Bautechnik 68, Heft 6, pp 206–212

Weber H (1993) Instandsetzung von feuchte- und salzgeschädigtem Mauerwerk.

expert-Verlag. Ehningen bei Böblingen

Weber WK (1999) Die gewölbte Eisenbahnbrücke mit einer Öffnung.

Begriffserklärungen, analytische Fassung der Umrisslinien und ein erweitertes

Hybridverfahren zur Berechnung der oberen Schranke ihrer

Grenztragfähigkeit, validiert durch einen Großversuch. Dissertation, Lehrstuhl

für Massivbau der Technischen Universität München

Weiss G (1992) Die Eis- und Salzkristallisation im Porenraum von Sandsteinen

und ihre Auswirkungen auf das Gefüge unter besonderer Berücksichtigung

gesteinsspezifischer Parameter. Dr. Friedrich Pfeil Verlag

Welch PJ (1995) Renovation of masonry bridges. Arch Bridges, C Melbourne

(Ed), Proceedings of the 1st International Conference on Arch Bridges,

Bolton, Thomas Telford, London, pp 601–611

Wenzel F (Ed) (1997): Mauerwerk – Untersuchen und Instandsetzen durch

Injizieren, Vernadeln und Vorspannen. Erhalten historisch bedeutsamer

264 6 Damages and Repair

Universität Karlsruhe

Wihr R (1980) Restaurierung von Steindenkmälern, Verlag Callwey

Witzany J, Čejka T & Zigler R (2008) Experimental research of masonry vaults

strengthening. In: Forde: Structural Faults and Repair 2008. 12th International

conference, 10–12th June 2008, Edinburgh

Woodward RJ (1997) Strengthening arch bridges. Annual Review 1997, TRL

Limited, Crowthorne

Zahn A (1999) Wieder ganz tragfähig – Kulturhistorisch wertvoller

Brückenbau gerettet und heutigen Verkehrsanforderungen angepasst.

Bautenschutz und Bausanierung. Köln 22, pp 29–30

7 Safety Assessment

the term “safety.” Often the term “safety” is defined as a situation with a

lower risk compared to an acceptable risk or as a situation “without any

impending danger.” Other definitions describe safety as “peace of mind.”

Whereas the first definition using the term “risk” is already based on a

substitution, the later term using “peace of mind” is a better definition. The

authors consider “safety” to be the result of an evaluation process of a cer-

tain situation. The evaluation can be carried out by every system that is

able to perform a decision-making process, such as animals, humans, so-

cieties, or computers that use some algorithms. However, algorithms usu-

ally use some numerical representation, such as risk R, for the description

of safety S:

existing R ≤ permitted R → S (7-1)

existing R > permitted R → S

In contrast, the authors consider not only numerical presentations as re-

sults of decision-making processes, but also human feelings. Therefore,

safety is understood here as a feeling. The decision-making process focus-

es mainly on preservation. Furthermore, the decision-making process deals

with whether some resources have to be spent to decrease hazards and

danger to an acceptable level. In other terms, “safety” is a feeling, which

describes that no further resources have to be spent to decrease any threats.

If one considers the term “no further resources have to be spent” as a de-

gree of freedom of resources, one can define “safety” as a value of a func-

tion that includes the degree of freedom of resources. Furthermore, one can

assume that the degree of freedom is related to some degree of distress and

relaxation. Whereas in safe conditions relaxation occurs, in dangerous sit-

uations a high degree of distress is clearly reached.

D. Proske, P. van Gelder, Safety of Historical Stone Arch Bridges, DOI 10.1007/978-3-540-77618-5_7,

© Springer-Verlag Berlin Heidelberg 2009

266 7 Safety Assessment

ranges from “danger” to “peace of mind,” and the value of the function as

degree of freedom of resources is shown in Fig. 7-1. It is assumed here that

the relationship is nonlinear, with at least one region of over proportional

growth of the relative freedom of resources. In Figure 7-1, this region of

over proportional growth is defined as the starting point of the safety

region:

S = { x | f ′′( x ) = 0} (7-2)

However, the question still remains: where does the region of safety

start since other points are possible? Such further points can be located

either at regions of maximum curvature or at the point of inflection.

manifested in many laws, like the human rights of the United Nations, the

German constitution with the right to life and personal integrity, the Prod-

uct Liability Act, civil code, or some administrative fiats. Codes of prac-

tice are administrative fiats and state the requirement that structures have

to be safe (Proske 2008b). Also, in the sense of building laws, structures

have to be safe and should not endanger public safety, life, and health. Es-

pecially in the codes, safety is understood as capability of structures to res-

ist loads. Reliability is then understood as a measure to provide this capa-

bility in different engineering fields. Here, a change from the general

qualitative statement to a quantitative statement becomes obvious. This is

very important for the engineers: now the engineer is enabled to prove

safety by computation. The reliability is meanly understood as probability

of failure (Fig. 7-2). Risk, which would be an alternative measure of

7.1 Definition of Safety and Safety Concepts 267

ween different accidental loads and emergency situations is possible

(Proske 2008b).

Fig. 7-2. Probability of failure for two random variables. First, (A) and (C) statis-

tical data about the load and the strength are investigated. Then, a statistical inves-

tigation is carried out (B) and (D). Both distribution functions resulting from the

statistical investigation are then merged to (E) further introducing a limit state

function g(X). In (F), the two-dimensional distribution function is shown in three-

dimensional illustrations of the probability of failure

for the exposure of indeterminacy and uncertainty. Other concepts, like

fuzzy-sets, rough-sets, Grey numbers, or further mathematical techniques

are not considered. However, research is carried out in this field. Figure

7-3 shows the different safety concepts for structures over time.

268

7 Safety Assessment

7.2 Probabilistic Safety Concept 269

tural engineering assumes that the exact value of many design variables is

unknown. This uncertainty is based on

• Random aberrations of characteristic values of the structural resistance

• Random aberrations by transferring laboratory test results to the struc-

ture

• Random aberrations of cross-section sizes and other geometrical meas-

ures

• Geometric imperfections

• Random aberrations of internal forces like moments, shear forces, or

axial forces

• Inherent uncertainties in the choice of characteristic value of loads

• Differences in the models for the loads

However, the stochastic models do not consider systematical errors like

computational errors in structural design processes or bad workmanship.

Such errors have to be avoided by control mechanisms (DIN 1055-100

1999).

7.2.1 Introduction

Mayer (1926) in Germany and Chocialov (1929) in the Soviet Union

(Murzewski 1974). In the third decade of the 20th century, the number of

people working in that field had already increased, just to mention

Streleckij (1935) in the Soviet Union, Wierzbicki (1936) in Poland, and

Prot (1936) in France (Murzewski 1974). Already in 1944 in the Soviet

Union, the introduction of the probabilistic safety concept for structures

had been forced by politicians (Tichý 1976). The development of probabil-

istic safety concepts in general experienced a strong impulse during and

after World War II, not only in the field of structures but also in the

field of aeronautics. In 1947, Freudenthal (1947) published his famous

work about the safety of structures. Until now, a model code for the pro-

babilistic safety concept of structures has been published by the JCSS (2004).

The probability of failure pf as proof measure for safety is computed as

function of the design values x. It can be referred to one year or the life-

time of the structures:

270 7 Safety Assessment

(7-3)

∫

p f = ... ∫

g (X) ≤ 0

f X ( x )dx

p f (n) = 1 − (1 − p f ) n . (7-4)

the probability of failure:

β = −Φ −1 ( pf ). (7-5)

Results are given in Table 7-1. The integration of the probability of fail-

ure volume can then be transferred into an optimization task to determine

the safety index. This is shown in Figs. 7-4 and 7-5.

The explained safety concept can be found in many regulations, such as

Eurocode 1 (1994), DIN 1055-100 (1999), GruSiBau (1981), and JCSS

Modelcode (2004). In these regulations, goal values for safety indexes can

also be found. These values are then the basis for the estimation of safety

factors, which are introduced for practical reasons.

Probability10–12 10–11 10–10 10–9 10–8 10–7 10–6 10–5 10–4 10–3 10–2 10–1 0.5

of failure

Safety 7.03 6.71 6.36 5.99 5.61 5.19 4.75 4.26 3.72 3.09 2.33 1.28 0.0

index

tion of the multidimensional probability may be simplified due to the low

value. The simplification explained in this section increases the speed of

the computation tremendously compared to a numerical integration of a

multidimensional space.

In general, the simplification is based on the transfer of the integration

of a multidimensional volume into an extreme value task. The result of this

extreme value computation is the substitute measure safety index. The

safety index itself describes the shortest distance between the origin in a

standard normal distributed space and the limit state function, g(X). A rela-

tionship between the probability of failure volume and the distance ex-

pressed by the safety index exists in this space (Table 7-1).

7.2 Probabilistic Safety Concept 271

tion with a mean value of 0 and a standard deviation of 1. The general as-

sumption of this procedure is the transformability of all arbitrary random

distribution functions into standard normal distribution functions. The sec-

ond assumption is the linearization of the limit state function. The lineari-

zation gave the following name for the technique: first-order reliability

method (FORM). The point of linearization on the limit state function is

the so-called design point. This point is characterized by maximum prob-

ability of failure at the limit state function.

Fig. 7-4. Transfer of the probability volume into an extreme value computation

However, a FORM computation not only delivers the safety index as re-

sult, but also measures which may be useful to compute partial safety fac-

tors, characteristic values, and design values. These values can be found in

many codes of practice and indicate the strong relationship between the

272 7 Safety Assessment

codes and this probabilistic safety concept. Therefore, current safety con-

cepts are called semiprobabilistic safety concepts.

In the following paragraphs, the FORM-methodology will be introduced

in detail. The method is often called the Rackwitz-Fießler (Fießler et al.

1976) algorithm or normal tail approximation. In general, the concept is

based on the fundamental work by Hasofer and Lind (1974).

In the procedure, first the non-normal distributed random variables have

to be transferred into normal random variables. The following formulas

will be used

1 ⎛ x i − m xi ⎞

* *

(7-6)

f xi ( x i ) = * ϕ ⎜

*

⎟

σ xi ⎜⎝ σ xi ⎟⎠

*

Fxi ( xi* ) = Φ ⎜ i

⎟

⎜ σ x* ⎟

⎝ i ⎠

with xi* as design point, m*xi as mean value, and σ*xi as standard devia-

tion of the normal distribution. Since the normal distribution should be

used as an approximation of the original distribution, mean value and stan-

dard deviation have to be computed by rearranging the formulas

1 (7-8)

σ x*i = *

ϕ (Φ −1 (Fxi ( xi* )))

f xi ( x i )

1. Define an iteration counter k = 0 and chose a design point for the first

iteration.

2. Transfer all non-normal distributed random variables into normal dis-

tributed random variables according to the following equations, with

i = 1, 2, …, m and m as number of random variables considered:

1 (7-10)

σ x*( k ) = ϕ (Φ −1 (Fxi ( xi( k ) )))

i

f xi ( xi( k ) )

7.2 Probabilistic Safety Concept 273

y (k)

= .

i

σ x*( k )

i

4. Compute the limit state function and the first derivative at yi( k ) :

∂h ∂g ∂xi ∂g (7-14)

= ⋅ = ⋅ σ*(i k ) .

∂y i y = y( k )

∂x i x = x( k )

∂yi ∂x i x = x( k )

point yi( k ) :

∂h (7-15)

∂y i y = y( k )

α(i k ) = 1/ 2

⎛ m ⎛ ⎞

2

⎞

⎜ ⎜ ∂h ⎟

⎜∑

⎟

⎟

⎜ j =1 ⎜⎝ ∂y j ⎟

⎠ ⎟

⎝ y=y

⎠

(k)

m

∂h (7-16)

h( y( k ) ) − ∑ y(jk )

j =1 ∂y j

y = y( k )

δ( k ) = 1/ 2

.

⎛ m ⎛ ⎞

2

⎞

⎜ ⎜ ∂h ⎟

⎜∑

⎟

⎟

⎜ j =1 ⎜⎝ ∂y j ⎟

⎠ ⎟

⎝ y=y

⎠

(k)

for i = 1, 2, … , m:

xi( k +1) = m*(x k ) − α i*( k ) ⋅ σ x*( k ) ⋅ δ ( k ) .

i i

(7-17)

7. Verify if xi( k +1) ≈ xi( k ) . If it is fulfilled, then the design point has been

found and the safety index is β = δ with h(0) > 0. If it is not fulfilled,

then the iteration starts again with Step 2.

This method is very practicable and has experienced a major spreading.

It gives fast and accurate results if the probability of failure is small, the

random distribution functions do not diverge too strong from the normal

distribution, and the limit state function does not show a strong curvature.

274 7 Safety Assessment

If the limit state function shows a strong curvature, then the curvature has

to be considered in the computation of the safety index (Fig. 7-6). This can

be done by the second-order reliability method. Here, the curvature of the

limit state function is approximated using

1 (7-18)

h ( y) = h ( y * ) + ( y − y* ) T ⋅ ∇ h ( y * ) + ( y − y * ) T ⋅ B y ⋅ ( y − y* ) = 0 .

2

By is the matrix of the second and mixed derivates from h(y) in the stan-

dardized space at the design point. Breitung (1984) has introduced the fol-

lowing equation with ai and i = 1, 2, … , m-1 as principal curvature of h at

the design point in the standard normal space

m −1 (7-19)

Pf = Φ(− β )∏ (1 − β ⋅ ai ) −1/ 2 .

i =1

jor part of this method. To compute the principal curvatures, a rotation of

the coordinate system is required. The rotation requires orthogonalizing

using the Schmidt process.

The old and new coordinates are linked as follows:

y = D⋅u . (7-20)

7.2 Probabilistic Safety Concept 275

u = DT ⋅ y . (7-21)

D = (d1 , d 2 ,..., d m )T (7-22)

with

d1 = α (7-23)

f

dk = k

fk

and

k −1 (7-24)

f k = e k − ∑ (eTk d l )d l

l =1

k = 2, 3, ..., m .

and ek is the kth unit vector. In the new system,

u* = DT y * = ( β ,0,0,...,0)T (7-25)

and

⎛ ∂g ⎞

T (7-26)

∇gu = D ⋅ ∇h = ⎜ u ,0,0,...,0 ⎟ .

⎝ ∂u1 ⎠

gu ( u * ) = 0 (7-27)

Bu = DT B y D (7-28)

∂gu 1 (7-29)

(u1 − u1* ) ⋅ + ( u − u * )T ⋅ B u ⋅ ( u − u * ) = 0 .

∂u1 2

The principal curvatures are the roots of the following equation:

⎛ B ˆ ⎞ (7-30)

det ⎜ u

− a⋅I⎟ = 0 ,

⎜ ∂g / ∂u ⎟

⎝ u 1 ⎠

276 7 Safety Assessment

When Bu

derived by deleting the first line and first column from Bu , I is the identity

matrix.

The approximation by Breitung gives good results, if the curvature of

the limit state function still is not too strong and the safety index is small.

Tvedt (1988) has introduced an extension of Breitung’s formulae:

γ = 1 − Pf (7-31)

γ = 1 − ( A1 + A2 + A3 ) (7-32)

with

n −1 (7-33)

A1 = Φ(− β )∏ (1 + β ⋅ a j ) −1/ 2

j =1

⎧⎪ n −1 n −1

⎫⎪ (7-34)

A2 = [ β ⋅ Φ(− β ) − ϕ ( β )] ⋅ ⎨∏ (1 + β ⋅ a j )−1/ 2 − ∏ (1 + ( β + 1) a j )−1/ 2 ⎬

⎩⎪ j =1 j =1 ⎭⎪

⎧⎪ n −1 ⎡ n −1 ⎤ ⎫⎪

⋅ ⎨∏ (1 + β ⋅ a j ) −1/ 2 − Re ⎢∏ (1 + ( β + i) ⋅ a j ) −1/ 2 ⎥ ⎬ .

⎪⎩ j =1 ⎣ j =1 ⎦ ⎪⎭

The extension of Breitung’s formula provides accurate results with low

and semi-low probabilities of failure. High probabilities of failure and

negative curvatures are not covered by the method. A further improvement

was done by the so-called Tvedt’s exact solution for parabolic limit state

function. Here, the size of the probability of failure is not limited:

γ = 1 − Pf (7-36)

1

∞

⎡ 1 n −1 ⎤ exp(−1/ 2θ 2 ) (7-37)

γ = 0.5 +

π ∫0 ⎣

sin ⎢ β ⋅ θ + ∑ arctan( a jθ ) ⎥ n −1 dθ .

2 j =1 ⎦ θ (1 + a 2θ 2 )1/ 4

∏j =1

j

integration.

7.2 Probabilistic Safety Concept 277

Additionally, Köylüoglu and Nielsen (1994) have tried to improve the ac-

curacy of SORM methods for higher probabilities of failure. They also

formulate

γ = 1 − Pf . (7-38)

n −1

1

γ = 1 − Φ (−β)∏

j =1 1 + a j / c0,1

⎧ ⎡⎛ n −1 ⎞

2

n −1 ⎛ ⎞ ⎤

2

⎪ 1 n −1 ak 1 ak ak

⋅ ⎨1 + c1,1 ∑ + c2,1 ⎜ ∑

⎢ + 2∑ ⎜ ⎥

2 1 + a / c 4 ⎢ ⎜ k =1 1 + a / c ⎟⎟ ⎜ 1 + a / c ⎟⎟ ⎥

⎪⎩ k =1 k 0,1

⎣⎝ k 0,1 ⎠ k =1 ⎝ k 0,1 ⎠

⎦

1 ⎡⎛ n −1 ⎞ ⎛ n −1 ⎛ ⎞ ⎞

3 2

ak ⎞ ⎛ n −1 ak ak

+ c3,1 ⎢⎜ ∑ ⎟ + 2 ⎜⎜ ∑ ⎟⎟ ∑ ⎜⎜

⎜ ⎟⎟ ⎟

8 ⎢⎜⎝ k =1 1 + ak / c0,1 ⎟⎠ ⎝ k =1 1 + a / c ⎜

0,1 ⎠ k =1 ⎝ 1 + a / c0,1 ⎠ ⎟

⎣ k

⎝ k

⎠

⎛ n −1 ⎞

3

⎤ ⎫

⎥ +…⎬⎪ .

ak

+12∑ ⎜ ⎟⎟

⎜

k =1 ⎝ 1 + ak / c0,1 ⎠ ⎥

⎦ ⎪⎭

(7-39)

If all curvatures ai are negative, then

n −1

1

γ = 1 − Φ(+β)∏

j =1 1 − a j / c0,2

⎧ n −1 ⎡⎛ n −1 ⎞

2

n −1 ⎛ ⎞

2

⎤

⎪ 1 ak 1 ak ak

⋅ ⎨1 + c1,2 ∑ + c2,2 ⎢⎜ ∑ ⎟⎟ + 2∑ ⎜⎜ ⎟⎟ ⎥

k =1 1 − ak / c0,2

⎜

⎢⎝ k =1 1 − ak / c0,2 k =1 ⎝ 1 + ak / c0,2 ⎥

⎪⎩ 2 4

⎣ ⎠ ⎠ ⎦

1 ⎡⎛ n −1 ak ⎞

3

⎛ n −1 ak ⎞ ⎛ n −1 ⎛ ak ⎞

2

⎞

+ c3,2 ⎢⎜ ∑ ⎟⎟ + 2 ⎜⎜ ∑ ⎟⎟ ⎜ ∑ ⎜⎜ ⎟⎟ ⎟

⎜

⎢⎝ k =1 1 + ak / c0,2

8

⎣ ⎠ ⎝ k =1 1 + ak / c0,2 ⎠ ⎝⎜ k =1 ⎝ 1 + ak / c0,2 ⎠ ⎟

⎠

⎛ n −1 ⎞

3

⎤ ⎫

⎥ + …⎪⎬ .

ak

+12∑ ⎜ ⎟⎟

⎜

k =1 ⎝ 1 + ak / c0,2 ⎥

⎠ ⎦ ⎪⎭

(7-40)

278 7 Safety Assessment

⎡ n −1 1 ⎧⎪ 1 n −1

ak ⎪⎫

(7-41)

γ = Φ( β ) + Φ( − β ) ⎢∏ ⎨1 + d1,2 ∑ + …⎬

⎢⎣ j = m 1 − a j / d0,2 ⎪⎩ 2 k = m 1 − ak / c0,2 ⎭⎪

⎧⎪ m −1 1 ⎪⎧ 1 m −1 ak ⎫⎫⎤

⎪⎪

⋅ ⎨1 − ∏ ⎨1 + c1,1 ∑ + …⎬⎬⎥

⎩⎪ j =1 1 + a j / c0,1 ⎪⎩ 2 k = m 1 + ak / c0,1 ⎪⎭⎭⎪⎦⎥

⎡ m −1 1 ⎧⎪ 1 m −1 ak ⎫⎪

− Φ ( β ) ⎢∏ ⎨1 + d1,1 ∑ + …⎬

k =1 1 + ak / c0,1

⎣⎢ j =1 1 + a j / d0,1 ⎩⎪ ⎪⎭

2

⎧⎪ n −1 1 ⎪⎧ 1 n −1

ak ⎫⎫⎤

⎪⎪

⋅ ⎨1 − ∏ ⎨ 1 + c1,2 ∑ + …⎬⎬⎥ .

⎪⎩ j = m 1 − a j / c0,2 ⎩⎪ 2 k =1 1 − ak / c0,2 ⎭⎪⎪⎭⎦⎥

Based on the cut-off of the terms, different approximation formulations

can be derived. For one term with positive curvature, the coefficient becomes

Φ( − β ) (7-42)

c0,1 =

ϕ (β )

and c1,1 = c2,1 = … = 0 . If two terms are chosen, then the coefficients can

be computed as

Φ( − β ) ⎛ 1 ⎞ (7-43)

c0,1 = ⎜⎜ ⎟

ϕ ( β ) ⎝ 1 + 1 − βΦ(− β ) / ϕ ( β ) ⎟⎠

ϕ (β ) βΦ(− β ) (7-44)

c1,1 = 1−

Φ(− β ) ϕ (β )

and c2,1 = c3,1 = … = 0 . With three terms, one achieves

1 ϕ (β ) (7-45)

− c1,1 =

c0,1 Φ( − β )

1 c βϕ ( β ) (7-46)

− 2 1,1 + 2c2,1 =

2

c0,1 c0,1 Φ(− β )

7.2 Probabilistic Safety Concept 279

− 3 + 6 = .

3

c0,1 2

c0,1 c0,1 Φ(− β )

one positive solution. The solution of c0,1 less than the value assessed with

the following formulae should be taken:

Φ(− β ) ⎛ 1 ⎞ (7-48)

c0,1 = ⎜⎜ ⎟

ϕ ( β ) ⎝ 1 + 1 − βΦ(− β ) / ϕ ( β ) ⎟⎠

Based on the cut off of the terms, different approximation formulations can

be derived. For one term with negative curvature, the coefficient becomes

Φ( β ) (7-49)

c0,2 =

ϕ (β )

and c1,2 = c2,2 = … = 0 . If two terms are chosen, then the coefficients

can be computed as

Φ( β ) ⎛ 1 ⎞ (7-50)

c0,2 = ⎜⎜ ⎟

ϕ ( β ) ⎝ 1 + 1 + βΦ( β ) / ϕ ( β ) ⎟⎠

−ϕ ( β ) βΦ( β ) (7-51)

c1,2 = 1+

Φ( β ) ϕ (β )

and c2,2 = c3,2 = … = 0 . With three terms, one achieves

1 ϕ (β ) (7-52)

+ c1,2 =

c0,2 Φ( β )

1 c βϕ ( β ) (7-53)

+ 2 1,2 + 2c2,2 = −

2

c0,2 c0,2 Φ( β )

+ 3 + 6 =

3

c0,2 2

c0,2 c0,2 Φ(− β )

with at least one positive solution. The solution of c0,2 less than the value

assessed with the following formulae should be taken.

280 7 Safety Assessment

Φ( β ) ⎛ 1 ⎞ (7-55)

c0,2 = ⎜⎜ ⎟

ϕ ( β ) ⎝ 1 + 1 − β Φ( β ) / ϕ ( β ) ⎟⎠

d0,2 = 2c0,2 d1,1 = d2,1 = … = 0 and d1,2 = d2,2 = … = 0 one-term formulation.

Cai and Elishakoff (1994) also try to improve Breitung’s method by ex-

tending the original formulae into a Taylor theorem. The application is

quite simple:

1 ⎛ β2 ⎞ (7-56)

Pf = Φ( β ) + exp ⎜ − ⎟ ( D1 + D2 + D3 + …) .

2π ⎝ 2 ⎠

The single elements of the Taylor theorem are

D1 = ∑ λ j (7-57)

j

1 ⎛ ⎞ (7-58)

D2 = − β ⎜ 3 ⋅ ∑ λ j2 + ∑ λ j λk ⎟

2 ⎝ j j ≠k ⎠

1 ⎛ ⎞ (7-59)

D3 = ( β 2 − 1) ⎜ 15 ⋅ ∑ λ j3 + 9 ⋅ ∑ λ j2 λk + ∑ λ j λk λl ⎟ .

6 ⎝ j j ≠k j ≠k ≠l ⎠

The basis for the computation of the elements is again the principal cur-

vatures. However, these values have already been computed, if Breitung’s

formulae were employed:

a j = −2 λ j . (7-60)

The latest SORM method was introduced by Zhao and Ono (1999a, b) and

by Polidori et al. (1999). The first method is based on the inverse fast Fou-

rier transformation (IFFT). These methods will not be explained here.

However, it should be noted that Zhao and Ono (1999a) not only give a

summary about the conditions under which the different introduced SORM

7.2 Probabilistic Safety Concept 281

FORM, when to use SORM methods, and when to use their own method

(Zhao and Ono 1999c). However, Breitung (2002) has criticized some as-

sumptions of the method by Zhao and Ono.

search for the minimum distance can also be carried out in spherical coor-

dinates (see Fig. 7-7). After transformation into spherical coordinates, the

angle and radius of a search vector are systematically adapted to find the

minimum distance.

known limit state function g(X). In other words, the limit state function

should be one formula. However, in many cases, such as the computation

of the ultimate load of an arch bridge, this requirement cannot be fulfilled.

For example, consider a finite element program code. In those cases, the

mathematical procedure has to be approximated by a surrogate, to reach to

an acceptable computation when the aforementioned techniques like

FORM and SORM are used. A procedure to develop such a surrogate is

the response surface methodology. In this methodology, a simplified more

282 7 Safety Assessment

tensive mathematical procedure originally established. The concept is easily

understood when the complicated mathematical computations are substituted

by some laboratory or field tests. There also, no function is known, but

should be introduced. Based on the test results, a function will be intro-

duced. Some general works about the concept can be found in Box and

Draper (1987) and applications in structural engineering are mentioned in

Bucher and Bourgund (1990) and Rajashekhar and Ellingwood (1993).

The concept can be described as follows. A certain function with the in-

put variables X and some functional constants K is given with

g = f ( X,K ) (7-61)

but can only be pointwise solved. Therefore, an approximation function

should be developed

g = f ( X,K ) . (7-62)

There are many different types of mathematical approximation func-

tions. Probably the most applied methods are quadratic functions, either

n n (7-63)

g~ = a + ∑ bi ⋅ xi + ∑ ci ⋅ xi2

i =1 i =1

g~ = A + XT ⋅ B + XT ⋅ C ⋅ X (7-64)

⎛ B1 ⎞ ⎛ C11 C12 … C1n ⎞ (7-65)

⎜ ⎟ ⎜ ⎟

⎜ B2 ⎟ ⎜ C21 ⎟.

B= , C=

⎜ ⎟ ⎜ ⎟

⎜⎜ ⎟⎟ ⎜⎜ ⎟

⎝ Bn ⎠ ⎝ Cn1 Cnn ⎟⎠

pending on the number of pointwise solutions, there is an under determined

number of solutions situation, an exact number, or an over determined

number of solutions to compute K. If an over determined number of solu-

tions exist, then some minimum error methods should be applied. For the

exact number of the solutions, the degrees of freedom of functions depend-

ing on the type of the function are shown in Table 7-2. Shapes of solutions

points are shown in Fig. 7-8.

7.2 Probabilistic Safety Concept 283

Table 7-2. Degrees of freedom for certain response surface functions (Weiland 2003)

Approximation function Degrees of freedom

g( x )

n

Linear regression g( x) = a + ∑ bi ⋅ xi n +1

i =1

n n n

Quadratic function g( x ) = a + ∑ bi ⋅ xi + ∑∑ cij ⋅ xi ⋅ x j 1

2 ⋅ (n + 1) ⋅ (n + 2)

including mixed terms i =1 i =1 j =1

n n

Quadratic function with- g( x ) = a + ∑ bi ⋅ xi + ∑ xi2 2n + 1

out mixed terms i =1 i =1

… 1

2 ⋅ (2 + 3n + 3n 2 )

with mixed terms

Polynomial third order

… 3n + 1

without mixed terms

Polynomial fourth order

… 1

2 ⋅ (2 + 3n + 5n2 )

with mixed terms

Polynomial fourth order

… 5n + 1

without mixed terms

Fig. 7-8. Two examples of chosen solution points for three variables (Weiland

2003)

The approximated response surface can then be updated after the next

probabilistic computation. This means that the response surface mainly

acts as a local approximation and does not give a good approximation over

the entire range of the original function. However, that is not required for

FORM/SORM computations:

284 7 Safety Assessment

g( x m( k ) ) (7-66)

x D( k=+m1) = x m( k ) + ( x Dk − x m( k ) )

( g( x m( k ) ) − g( x Dk ))

with xm as centre point, xD as design point based on a FORM computa-

tion using the response surface, and k as iteration counter. The iteration

scheme is shown in Fig. 7-9.

Fig. 7-9. Iterative improvement of the response surface (Klingmüller and Bourgund

1992)

schema can easily be extended to existing finite element programs or other

numerical tools. The computation is easily understandable and the number

of computations is low.

The major disadvantage is a limited capability to find the extreme value

of complicated functions. The number of iterations also depends on the

number of random variables. Therefore, in high-dimensional cases, the re-

sponse surface method may perhaps cause heavy computations. Further-

more, the approximation method only uses data points from one iteration

cycle. However, it may perhaps be useful to keep the data for further in-

vestigations. There exist external programs that can carry out response sur-

face computations afterwards by using all available data.

Since the limitations of the response surface method are known, in the

last few years many new methods have been developed (Roos and Bayer

2008, Ross and Bucher 2003). An adaptive response surface method has

been introduced by Most (2008).

7.2 Probabilistic Safety Concept 285

In contrast to the FORM and SORM methods, the Monte Carlo Simulation

is an integration procedure, not an extreme value computation. Therefore,

some assumptions required for the FORM/SORM are not relevant for the

Monte Carlo Simulation. Also, Monte Carlo Simulation is extremely easy

to program and to apply. However, if the probabilities of failure are ex-

tremely low and as required by codes, then Monte Carlo Simulation will

require extensive computation power.

The general idea of Monte Carlo Simulation is, as the name already in-

dicates, the application of pure random numbers into a computation flow.

In its simplest description, Monte Carlo Simulation is an extensive version

of trial and error. The only assumption for this technique is some quality

requirements for random numbers. Since computers cannot provide real

random numbers, they produce pseudo-random numbers based on purely

deterministic causal computations; the period of the numbers should be big

enough so that random numbers are not repeated in the Monte Carlo Simu-

lation. There are many programs available to provide high-quality random

numbers (NR 1992). The Monte Carlo Simulation is then

2 (7-68)

f2 − f

∫f dV ≈ V f ± V

N

,

where V represents the volume, V f stands for the mean value of the

function f over the sample size N, and the ±-term gives more or less an one

standard deviation error estimator. The further functions are

N N

1 1 (7-69)

f ≡

N

∑ f ( xi )

i =1

f2 ≡

N

∑f

i =1

2

( xi ) .

The major advantage of the Monte Carlo Simulation concerning the di-

mensions is the fact that the statistical error is independent from the num-

ber of dimensions. For Monte Carlo Simulation, it does not matter if there

is 1 or 50 random input variables, the statistical error will remain the same.

This is completely different for Simpson’s rule applied for integration,

where the required computation grows exponentially with the number of

dimensions.

286 7 Safety Assessment

However, the required sample size has a major influence on the error

size in Monte Carlo Simulation. The next equation is an example to evalu-

ate the required sample size nreq (Flederer 2001)

1 (1 − Pf ) (7-70)

nreq = ⋅

1 − Pε Pf ⋅ ε 2

tistical error. From the equation, it becomes understandable that with low

probabilities of failure and low statistical error, high sample sizes are re-

quired. Macke (2000) has given a good example, where the probability of

–6

failure was 10 and the required statistical error was less than 50%. The

6

required sample size was 4 × 10 . If the statistical error should be less than

8

10%, then more than 10 samples are required. Based on these properties,

Monte Carlo Simulation is a good method for high probabilities of failure

and for high dimensions. If lower probabilities were to be investigated

with Monte Carlo Simulation, the so-called variance reduction techniques

should be applied. This is mainly done after a preliminary FORM/SORM

computation.

Since a great variety of variance reduction techniques exist, here only a

few will be mentioned. Probably the most applied technique is importance

sampling. Here, the original probability integral

Pf = ∫ ... ∫ f x (x)dx (7-71)

g<0

Pf = ∫ ... ∫ I ( x ) f x (x)dx . (7-72)

all

⎧ 1 g( x ) < 0 (7-73)

I ( x) = ⎨

⎩ 0 g( x ) ≥ 0

This permits a transformation of the random distributions using a chosen

distribution function hv(v):

f x (v ) (7-74)

Pf = ∫ ... ∫ I (v) hv (v)dv .

all

hv (v)

7.2 Probabilistic Safety Concept 287

1 mc

f x (v n ) (7-75)

Pˆ f =

mc

∑ I (v n )

n =1 hv (v n )

.

1 ⎡1 m c⎛ f ( v ) ⎞

2

⎤ (7-76)

Var[ Pˆ f ] = ⎢ ∑ I (v n ) ⎜ x n

⎟ − Pˆ f ⎥ .

2

mc − 1 ⎢ mc n =1 ⎝ hv (v n ) ⎠ ⎥⎦

⎣

The major task in applying importance sampling is the search for a

proper distribution function hv(v). If some prior information is available,

for example by FORM/SORM computation, then the distribution function

hv(v) can be selected very efficiently, yielding to an impressive drop of

computational effort in the Monte Carlo Simulation. In general, impor-

tance sampling can be understood as a transformation of the random points

toward interesting regions in the sampling space (Maes et al. 1993, Song

1997, and Ibrahim 1991).

The concept can be even further extended by updating the distribution

function hv(v) after every sample step. This technique is called adaptive

sampling (Bucher 1988 and Mori and Ellingwood 1993).

Another interesting technique is the application of quasi-random numbers

instead of pseudo-random numbers. Quasi-random numbers are not ran-

dom at all. They are constructed to fill a multidimensional space in a most

efficient way. Also, they can be understood as a technique standing be-

tween the classical Simpson’s rule for integration and the crude Monte

Carlo Simulation. Since the numbers are deterministic, the computation

error becomes related to the dimensions.

However, the major advantage of the application of quasi-random num-

bers with Monte Carlo Simulation is the fact that they can be applied to

many finished programs. So, if Monte Carlo Simulation was chosen for a

certain project and it turns out after some sample computations that the

computation time is unacceptable, but the programming cannot be changed

anymore, then perhaps quasi-random numbers can be produced externally

and can then given to the program. This technique was applied in Flederer

(2001). More details about possible application can be found in Curbach

et al. (2002).

288 7 Safety Assessment

Whereas for Monte Carlo Simulation the number of limit state functions is

irrelevant for the FORM and SORM methods, perhaps different limit state

functions were considered separately and have to be merged into one sin-

gle probability of failure or safety index. This can be seen at an arch

bridge, where several different point of failure can be identified. The ques-

tion that follows is whether these points are correlated or not.

If the different limit state functions are uncorrelated, then system prob-

ability of failure can be computed as

n (7-77)

Pf = 1 − ∏ (1 − Pfj ) .

j =1

If the single probabilities of failure are rather small, the values can be

added instead of multiplying

n (7-78)

Pf ≈ ∑ Pfj .

j =1

1⎛ n ⎞

2 (7-79)

E ( Pf ) ≤ ⎜ ∑ Pfj ⎟ .

2 ⎝ j =1 ⎠

If the probabilities of failure are expressed as safety indexes, the formu-

las become

⎛ n

⎞ ⎛ n

⎞ (7-80)

β sys = −Φ −1 ⎜1 − ∏ Φ( β j ) ⎟ ≈ −Φ −1 ⎜ ∑ Φ( − β j ) ⎟ .

⎝ j =1 ⎠ ⎝ j =1 ⎠

If the noncorrelation of the limit states does not hold true, then the cor-

relations have to be considered and a multidimensional normal distribution

can be applied. The correlations are expressed as

ρ jk = α Tj α k = α j1α k1 + α j 2α k 2 + ... + α jmα km (7-81)

and

α Tj = (α j1 + α j 2 + ... + α jm ) . (7-82)

With α Tj are the weighting factors of the m random variables from the

j limit state function giving the safety index βj. For the correlation matrix

th

7.2 Probabilistic Safety Concept 289

⎜ ⎟

ρ 1

R = ⎜ 21 ⎟

⎜ ⎟

⎜⎜ ρ ρn2

⎟

1 ⎟⎠

⎝ n1

and a vector for the safety indexes

⎛ β1 ⎞ (7-84)

⎜ ⎟

β

β = ⎜ 2 ⎟.

⎜ ⎟

⎜⎜ ⎟⎟

⎝ βn ⎠

The system probability of failure is obtained by transformation in the

standard normal space and linearization of the limit state functions:

Pf = 1 − Ps (7-85)

⎛ n ⎞ (7-86)

= 1 − P ⎜ ∩ ( g j ( X ) ≥ 0) ⎟

⎝ j =1 ⎠

⎛ n ⎞ (7-87)

= 1 − P ⎜ ∩ (h j (Y ) ≥ 0) ⎟

⎝ j =1 ⎠

⎛ n ⎞ (7-88)

≈ 1 − P ⎜ ∩ (l j (Y ) ≥ 0) ⎟

⎝ j =1 ⎠

⎛ n ⎞ (7-89)

= 1 − P ⎜ ∩ (− Z *j ≤ β j ) ⎟

⎝ j =1 ⎠

= 1 − Φ n (β , R) , (7-90)

βn β1

1 ⎛ 1 ⎞

Φn ( β , R ) = ∫ ∫ exp ⎜⎝ − 2 ψ R −1y ⎟ dy1 ,dy2 ,

T

1/ 2

dy n .

(2π ) n/2

R −∞ −∞ ⎠

(7-91)

290 7 Safety Assessment

Several different programs can be found for the evaluation of this function

(Schervish 1984, Genz 1992, Drezner 1992, and Yuan and Pandey 2006).

To simplify the computation, it is often assumed that the correlations are

equal between different limit state functions. The correlation matrix becomes

⎛1 ρ ρ⎞ (7-92)

⎜ ⎟

ρ 1

R=⎜ ⎟

⎜ ⎟

⎜ ⎟

⎝ρ ρ 1⎠

and the multidimensional standard normal distribution changes to

+∞ n ⎛ βi + ρ ⋅ x ⎞ (7-93)

Φ n ( β , R) = ∫ ϕ ( x)∏ Φ ⎜ ⎜ ⎟ dx .

1 − ρ ⎟⎠

−∞ i =1 ⎝

Based on this idea of a mean correlation matrix, the coefficients of the

correlation matrix can be understood as products of single elements, like

⎛ 1 λ1 ⋅ λ2 λ1 ⋅ λn ⎞ (7-94)

⎜ ⎟

λ ⋅λ 1

R=⎜ 2 1 ⎟ , λ < 1, λ < 1, i, j = 1, 2, ..., n

⎜ ⎟ i i

⎜⎜ ⎟⎟

⎝ λn ⋅ λ1 λn ⋅ λ2 1 ⎠

and the multidimensional standard normal distribution again changes

+∞ n ⎛β +λ ⋅x⎞ (7-95)

Φ n (β , R) = ∫ ϕ ( x)∏ Φ ⎜ ⎟ dx .

i i

⎜ 1 − λ2 ⎟

−∞ i =1

⎝ i ⎠

Suggestions of lower bounds for the correlations values are given as

λ j ⋅ λk ≤ ρ jk j ≠ k (7-96)

λ j ⋅ λk ≥ ρ jk j ≠ k (7-97)

and

− 1 ≤ λ j , λk ≤ 1 (7-98)

7.2 Probabilistic Safety Concept 291

λ j ⋅ λk = ρ jk . (7-99)

That will only be possible in some rare cases. However, if the ρjk has

approximately the same size and is positive, then the following recom-

mendation for lower bounds

λ j = max{ρ jk } j ≠ k (7-100)

j

λ j = min{ρ jk } j ≠ k . (7-101)

j

can be given.

If the contribution of the limit state functions to the system probability

of failure can be roughly estimated, then the three limit state functions with

the highest probability of failure or the lowest safety index can be chosen,

and the following computation can be carried out:

λ1 ⋅ λ2 = ρ12 , λ2 ⋅ λ3 = ρ 23 , λ1 ⋅ λ3 = ρ13 (7-102)

λ1 = , λ2 = , λ3 = .

ρ23 ρ13 ρ12

For the remaining values, an upper bound

⎧ ρ jk ⎫ (7-104)

λ j = min ⎨ ⎬ j = 4,5...

k ≤ j +1 λ

⎩ k ⎭

and a lower bound can be estimated

⎧ ρ jk ⎫ (7-105)

λ j = max ⎨ ⎬ j = 4,5... .

k ≤ j +1

⎩ λk ⎭

Unfortunately, the requirement

−1 ≤ λ j , λk ≤ 1 (7-106)

For series system, further bounds can be given. In serious systems,

the probability of failure increases by decreasing the correlation bet-

ween the different elements or limit state functions. This can be understood

292 7 Safety Assessment

written as

n n (7-107)

max Pfj ≤ Pf ≤ 1 − ∏ (1 − Pfj ) < ∑ Pfj

j

j =1 j =1

⎛ n

⎞ ⎛ n ⎞ (7-108)

min β j ≥ β sys ≥ −Φ −1 ⎜ 1 − ∏ Φ( β j ) ⎟ > −Φ −1 ⎜ ∑ Φ(− β j ) ⎟ .

j

⎝ j =1 ⎠ ⎝ j =1 ⎠

Based on general additive theorems for probabilities, more precise

bounds can be given:

⎧1 (7-109)

⎪

Pf ≤ min ⎨ n n

⎪ ∑ P ( Fj ) − ∑ max P(Fj ∩ Fk )

k< j

⎩ j =1 j =2

⎧0 (7-110)

⎪n

Pf ≥ P( F1 ) + ∑ max ⎨ j −1

j =2 ⎪

⎩

P ( Fj ) − ∑

k =1

P (Fj ∩ Fk ) .

P ( Fj ∩ Fk ) ≈ Φ 2 ( − β j , − β k ; ρ jk ) . (7-111)

⎧1 (7-112)

⎪

Pf ≤ min ⎨ n n

⎪ ∑ Φ ( − β j ) − ∑ max Φ 2 (− β j , − β k ; ρ jk )

k< j

⎩ j =1 j =2

⎧0 (7-113)

⎪n

Pf ≥ Φ(− β1 ) + ∑ max ⎨ j −1

j =2 ⎪Φ(− β j ) − ∑ Φ 2 ( − β j , − β k ; ρ jk ) .

⎩ k =1

with

7.2 Probabilistic Safety Concept 293

+∞ ⎛ x + λ ⋅w ⎞ ⎛ x + λ ⋅w ⎞ (7-114)

Φ 2 ( x2 ; x2 ; ρ ) = ∫ Φ ⎜ 1 1 ⎟ ⋅ Φ ⎜ 1 1 ⎟ ⋅ ϕ (w) dw

⎜ 1 − λ2 ⎟ ⎜ 1 − λ2 ⎟

−∞ ⎝ 1 ⎠ ⎝ 1 ⎠

using

λ1 = λ2 = ρ , ρ > 0 (7-115)

λ1 = − ρ , λ2 = − − ρ , ρ < 0 .

The introduced methods have been programmed into FORTRAN 77

routines and are available free of charge to the reader of this book. Please

simply contact the authors.

Arch bridges are usually considered to be a serious system: if one part

fails, the entire bridge will collapse. However, some parts of bridge behave

like parallel systems: one part will collapse and other parts will take more

loads. Further works about the estimation of probabilities of failure for

such types of systems can be found in Rackwitz and Hohenbichler (1981)

or Gollwitzer and Rackwitz (1990), as seen in Fig. 7-10.

Fig. 7-10. Relationship between system safety index and material properties

shown for a Daniel system

loads by random distribution functions. They have not considered any

294 7 Safety Assessment

work. For the interested reader, the Rosenblatt transformation or the

NATAF transformations are dealt with in Melchers (1999) or Liu and Der

Kiureghian (1986). Copulas are an additional technique to transform the

correlated random variables into noncorrelated variables.

Furthermore, these random distribution functions are usually kept con-

stant over different distances or volumes. In contrast, it is well-known that,

for example, material properties might not have a full correlation over a

certain distance or volume. This correlation change over distance may be

described by random fields (Vanmarcke 1983). In past years, random

fields are increasingly applied in structural safety investigations to estab-

lish stochastic finite elements (Der Kiureghian and Ke 1988, Ghanem and

Spanos 1991, and Pukl et al. 2006). The latest advances were shown by

Bayer and Ross (2008).

Furthermore, not all presented techniques perform well under all condi-

tions. Figures 7-11 and 7-12 give a good overview about the application

conditions of the probabilistic techniques.

It is not intended by the authors to give here a full summary about cur-

rent state of knowledge in the field of structural safety. A state-of-the-art

report for computational stochastic mechanics was given by Berman et al.

(1997), however, the latest developments should be considered. In con-

trast, the introduced methods can be easily programmed and applied to

arch bridge problems by the reader. For more advanced studies, some com-

mercial programs or programs from research institutes can be used.

Fig. 7-11. Performance of methods for stochastic structural analysis (Bucher et al.

2000)

7.2 Probabilistic Safety Concept 295

random variables and the estimated probability of failure (Bayer 2008)

of probabilities of failure, in many cases these programs lack a sufficient

manual, a graphical user interface, or simply an easy handling. Therefore,

in many cases, commercial probabilistic programs were developed.

Currently, the following programs are known to the authors, however

this list is subject to change: UNIPASS (Lin and Khalessi 2006), ProFES

(Wu et al. 2006), Proban (Tvedt 2006), PHIMECA (Lemaire and Pendola

2006), PERMAS-RA/STRUREL (Gollwitzer et al. 2006), NESSUS

(Thacker et al. 2006), COSSAN (Schueller and Pradlwarter 2006), CalRel/

FERUM/ OpenSees (Der Kiureghian et al. 2006), ANSYS PDS und

DesignXplorer (Reh et al. 2006), ATENA/SARA/FREET (Pukl et al.

2006), VaP (Petschacher 1994), OptiSlang (Schlegel and Will 2007),

RELSYS (Estes and Frangopol 1998), and the probabilistic toolbox

ProBox (Schweckendiek and Courage 2006). Many of these programs can

be downloaded free of charge for test runs (Table 7-3).

296 7 Safety Assessment

further stand alone programs were or are under development in companies

or at universities. Therefore, Epstein et al. (2008) suggested some general

requirements and structures for probabilistic computer programs. The ad-

aptation of these rules will ease the application of such programs and will

extend the user group.

Even without commercial programs, simple FORM or Monte Carlo

Simulations can be carried out with standard spreadsheet software. Includ-

ing an optimization tool like the solver in EXCEL, it is possible to com-

pute the safety index. An example of such an application can be found in

Low and Teh (2000).

Program University Homepage

VAP ETH Zurich http://www.ibk.baum.ethz.ch/proserv/vap.html

CALREL University of http://www.ce.berkeley.edu

California,

Berkeley

FERUM University of http://www.ce.berkeley.edu/~haukaas/FERUM/ferum.

California, html

Berkeley

NESSUS Southwest http://www.nessus.swri.org/

Research Institute,

San Antonia

PERMAS INTES GmbH, http://www.intes.de

Stuttgart

SLANG Bauhaus http://www.uni-weimar.de/Bauing/ism/Slang

University,

Weimar

ISPUD University http://www.uibk.ac.at/c/c8/c810

Innsbruck

COSSAN University http://www.uibk.ac.at/c/c8/c810

Innsbruck

PROBAN Det Norske http://www.dnv.com

Veritas Software

STRUREL RCP GmbH http://www.strurel.de

ANSYS ANSYS Inc. http://www.ansys.com

SARA University Bruno,

FREET Cervenka http://www.cervenka.cz

Consulting

RACKV University of

Natural Resources

and Applied Life

Sciences, Vienna

7.2 Probabilistic Safety Concept 297

of the structure, they have to be compared to a certain proof or goal value.

Many references have published such goal probabilities of failure and

examples are shown in Tables 7-4 to 7-10. However, most of the publica-

tions show nearly the same values; for new structures, maximum probabil-

–6

ity of failure is in the range of 10 per year or a minimum safety index of

3.8 per year. For existing structures, usually less stringent requirements are

common. For example, a decrease of the safety index by 0.5 (Diamantidis

et al. 2007). Furthermore, the probability of failure or the safety index can

be adapted to more special conditions.

Table 7-4. Goal probability of failure per year according to the CEB (1976)

Average number of people endangered Economical consequences

low average high

–3

Low ( < 0.1) 10 10–4 10–5

–4 –5

Average 10 10 10–6

–5 –6

High ( > 10) 10 10 10–7

1992)

Safety Failure consequences Probability of failure for the limit

class state of ultimate load per year

Low Low personal injuries 1.0 ×·10–4

Insignificant economical consequences

Normal Some personal injuries 1.0 × 10–5

Considerable economical consequences

High Considerable personal injuries 1.0 × 10–6

Very high economical consequences

Table 7-6. Goal probability of failures in the former East Germany (Franz et al.

1991)

Reliability class Consequences Probability of failure

I Very high danger to the public 1.0 × 10–7

Very high economical consequences

Disaster

II High danger to the public 1.0 × 10–6

High economical consequences

High cultural losses

III Danger to some persons 1.0 × 10–5

Economical consequences

298 7 Safety Assessment

IV Low danger to persons 1.0 × 10–4

Low economical consequences

V Very low danger to persons 7.0 × 10–4

Very low economical consequences

Safety Possible consequences of failure Type of limit state

class

Limit state of ultimate Limit state of Ultimate Serviceability

load bearing serviceability load

1 No danger to humans Low economical 1.34 × 10–5 6.21 × 10–3

and no economical consequences and

consequences low usage limitation

2 Some danger to humans Considerable 1.30 × 10–6 1.35 × 10–3

and considerable economical

economical consequences and

consequences strong limitation of

further usage

3 High importance of the High economical 1.00 × 10–7 2.33 × 10–4

structure to the public consequences and

high restriction to

future usage

Table 7-8. Goal probability of failure according to the DIN 1055-100 (1999) and

the Eurocode 1 (1994)

Limit state Probability of failure

Lifetime Per year

–5

Ultimate load 7.24 × 10 1.30 × 10–6

–2

Serviceability 6.68 × 10 1.35 × 10–5

Limit state Safety index

Serviceability

Reversible 0.0

Irreversible 1.5

Fatigue

Testable 2.3

Not testable 3.1

Ultimate load

Very low consequences 2.3

Low consequences 3.1

Common consequences 3.8

High consequences 4.3

7.2 Probabilistic Safety Concept 299

Table 7-10. Goal safety indices according to the JCSS Modelcode (2004)

Costs for safety Low failure Average failure High failure

measures consequences consequences consequences

Low 3.1 3.3 3.7

Average 3.7 4.2 4.4

High 4.2 4.4 4.7

consequence classes in terms of failure consequence classes (CC) as

shown in Table 7-11. Such consequence classes can then be related to

some reliability classes (RC) listed in Tables 7-12 and 7-13.

Failure CCs Consequences Examples

CC 3 High consequences to Stands, public buildings, for

humans, the economy, example concert halls

social systems and the

environment

CC 2 Average consequences to Dwelling and office buildings,

humans, the economy, public buildings such as offices

social systems and the

environment

CC 1 Low consequences to Agricultural structures or

humans, the economy, structures without regular

social systems, and the persons’ residence, for example

environment barns, conservatories

RC Safety index per year Safety index for 50 years

RC 3 5.2 4.3

RC 2 4.7 3.8

RC 1 4.2 3.3

300 7 Safety Assessment

Table 7-13. Adaptation factor for the partial safety index subject to the RC (Euro-

code 1 1994)

Adaptation for the par- RC

tial safety factors RC1 RC2 RC3

KFI 0.9 1.0 1.1

control of the building material in terms of changes of partial safety factors

of the material. Still, this is for new structures only. Therefore, some other

recommendations focus on existing structures. For example, in Tables 7-14

and 7-15, some adaptation factors are given. Furthermore, Strauss and

Bergmeister (2005) have also introduced some factors.

Canadian Limit States Design Standard (taken from Casas et al. 2001 and COST

345 2004)

β = 3.5 − ( Δ E + Δ S + Δ I + Δ PC ) ≥ 2.0 value

Correction factor for element failure ΔE

Abrupt failure without warning 0.0

Abrupt loss of bearing capacity without warning with remaining capacity 0.25

Grateful failure with warning 0.50

Correction factor for system failure ΔS

Failure of one single element causes system failure 0.00

Failure of one single element does not cause system failure 0.25

Failure of one single element causes local failure only 0.50

Correction factor for monitoring ΔI

Element is not controllable –0.25

Element is controlled regularly 0.00

Critical elements are controlled more frequently 0.25

Correction factor for live load Δ PC

All types of traffic without special permission 0.00

All types of traffic with special permission 0.60

Table 7-15. Adaptation of the safety index according to Schueremans and Van

Gemert (2001)

β = βT − ( Δ S + Δ R + Δ P + Δ I ) ≥ 2.0 value

Adjustment for system behaviour ΔS

Failure leads to collapse, likely to impact occupants 0.00

Failure is unlikely to lead to collapse, or unlikely to impact occupants 0.25

Failure is local only, very unlikely to impact on occupants 0.50

Adjustment for risk category ΔR

High number of occupants (n) exposed to failure (n = 100–1,000) 0.00

7.3 Semiprobabilistic Safety Concept 301

β = βT − ( Δ S + Δ R + Δ P + Δ I ) ≥ 2.0 value

Normal occupancy exposed to failure (n = 10–99) 0.25

Low occupancy exposed to failure (n = 0–9) 0.50

Adjustment for past performance: ΔP

No record of satisfactory past performance 0.00

Satisfactory past performance or dead load measured 0.25

Adjustment for inspection: ΔI

Component not inspect able –0.25

Component regularly inspected 0.00

Critical component inspected by expert 0.25

This adapted safety index can then be used to provide alternative safety

measures in the semiprobabilistic safety concept.

Further background information about the development of goal prob-

abilities of failure or safety indexes can be found in Proske (2008b) dis-

cussing different risk parameters and the current developments.

7.3.1 Introduction

The probabilistic safety concept for structures has introduced the measure

of probability of failure as a measure of reliability and safety. Neverthe-

less, under everyday conditions, this measure is not practical and instead,

simpler types of proof of safety have to be used for structures. Therefore,

the probabilistic safety concept has to be transformed into a semiprob-

abilistic safety concept, which means nothing else than developing substi-

tutes for the probability of failure proof, which are easier to handle. Such

substitutes are the safety factors, characteristic values, and the design val-

ues. They are developed on some basic simplification.

One major advantage for these elements is the long tradition of the ap-

plication of safety factors. It has been estimated that the first application of

a global safety factor goes back up to 300 B.C. by Philo from Byzantium

(Shigley and Mischke 2001). He introduced the global safety factor in

terms of

resistance (1-116)

γ= .

load

302 7 Safety Assessment

Empirical geometrical rules remained valid over nearly the next two

millenniums. Only in the last few centuries have the applications of safety

factors become widespread. Over time, several different values were de-

veloped for different materials. In most cases, the values dropped signifi-

cantly during the last century. As an example, in 1880 for brick masonry

the safety factor of 10 was required, whereas only 10 years later the factor

was chosen between 7 and 8. In the 20th century, the values ranged from

factor 5 and 4, and now for the recalculation of historical structures with that

material, factor 3 is chosen (Busch and Zumpe 1995, Schleicher 1949,

Wenzel 1997, Mann 1987, and Tonon and Tonon 2006). This decline of

safety factors could also be observed for other materials such as steel. The

development of new materials especially led to more concerns about the safe

application of those materials. The different developments for safety factors

for materials led to the first efforts in the beginning of the 20th century to

develop material-independent factors, as shown in Tables 7-16 and 7-17.

Safety factor Knowledge of load Knowledge of Knowledge of

material environment

1.2–1.5 Excellent Excellent Controlled

1.5–2.0 Good Good Constant

2.0–2.5 Good Good Normal

2.5–3.0 Average Average Normal

3.0–4.0 Average Average Normal

3.0–4.0 Low Low Unknown

Safety Knowledge of load Knowledge of Knowledge of

factor material environment

1.3 Extremely well-known Extremely Likewise tests

well-known

2 Good approximation Good approximation Controllable

environment

3 Normal approximation Normal Moderate

approximation

5 Guessing Guessing Extreme

limited; otherwise, the major requirement safety of structures might not be

fulfilled anymore. Therefore, more advanced changes might be considered

to meet the demanding requirements of economic and safe structures. Such

7.3 Semiprobabilistic Safety Concept 303

the table 7-16; for example, a safety factor for the load and a safety factor

for the material. This is indeed the idea of the partial safety factor concept.

It does not necessarily yield to lower safety factors, but it yields to a more

homogenous level of safety.

The development of partial safety factors is strongly related to the de-

velopment of the probabilistic safety concept. However, the practical appli-

cation of partial safety factors took additional decades. First applications

can be found in steel design, whereas a first example in the field of struc-

tural concrete could be ETV concrete—a concrete code in East Germany

(ETV stands for Unified technical codes). ETV concrete was developed

during the seventies of the 20th century and introduced in the beginning of

the eighties. A comparison between ETV and the up-to-date German code

DIN 1045-1 can be found in Wiese et al. (2005).

7.3.2.1 Introduction

The change from the global safety factor concept to the partial safety fac-

tor concept can be easily seen in the following equations. In general, the

comparison of the resistance of a structure and the load or event remains

E d ≤ Rd . (7-117)

But in contrast to the global safety factor format, where the safety factor

can be separated like

Ed ≤ Rd / γ Global , (7-118)

the safety factors accompany the parameters required for design. Then,

the design load Ed is evaluated according to

Ed = ∑ γ G , j ⋅ Gk , j + γ Q ,1 ⋅ Qk ,1 + ∑ γ Q ,i ⋅ψ 0,i ⋅ Qk ,i (7-119)

j ≥1 i >1

⎛ f f yk f f p 0,1k f pk ⎞ (7-120)

Rd = R ⎜ α ⋅ ck ; ; tk ,cal ; ; ⎟.

⎝ γc γs γs γs γs ⎠

A list of material partial safety factors is given in Table 7-18.

304 7 Safety Assessment

Accidental

Code or Limit state of

Material load

reference ultimate load

conditions

Concrete (up to C 50/60) DIN 1045-1 1.50 1.30

1.5/ 1.3/

Concrete higher than C 50/60 DIN 1045-1

(1.1-fck/500) (1.1-fck/500)

Non-reinforced concrete DIN 1045-1 1.80 1.55

Non-reinforced concrete DAfSt (1996) 1.25

Pre-cast concrete DIN 1045-1 1.35

Collateral evasion DIN 1045-1 2.00 –

Reinforcement steel DIN 1045-1 1.15 1.00

Pre-stressing steel DIN 1045-1 1.15 1.00

Steel yield strength EN 4 1.10 1.00

Steel tensile strength EN 4 1.25 1.00

Steel tensile strength EN 4 1.00 1.00

Wood EN 5 1.30 1.00

Masonry (Category A) EN 6 1.7 (I)/2.0 (II) 1.20

Masonry (Category B) EN 6 2.2 (I)/2.5 (II) 1.50

Masonry (Category C) EN 6 2.7 (I)/3.0 (II) 1.80

Masonry – steel EN 6 1.50/2.20

Masonry DIN 1053-100 1.50–1.875 1.30–1.625

Wall anchorage (C. A-C) Mann (1999) 2.50 1.20

Floatglas/Gussglas BÜV (2001) 1.80 1.40

ESV-Glas BÜV (2001) 1.50 1.30

Siliconglas BÜV (2001) 5.00 2.50

Carbon fibre Onken et al. 1.20

Carbon fibre (2002) and 1.301

Carbon fibre cable Bergmeister 1.201

Carbon fibre glue (2003) 1.501

Cladding DIN 18 516 2.00

Aluminum yield strength EN 9 1.10

Aluminum tensile strength EN 9 1.25

Bamboo as building material Bamboo (2005) 1.50

Textile reinforced concrete Own works 1.802

Concrete – multiaxial loading Own works 1.35–1.70

Granodiorit tensile strength Own works 1.50–1.70

Historical masonry arch bridges UIC-Codex 2.00 3

Historical non-reinforced arch Bothe et al. 1.803

bridges (2004)

1

Consider construction conditions (Bergmeister 2003).

2

Recent researches indicate lower values.

3

A partial safety factor for a system.

7.3 Semiprobabilistic Safety Concept 305

Unfortunately, the definition of the safety format for the resistance de-

pends on the way of computing the forces in the structure. Eibl (1992) and

Eibl and Schmidt-Hurtienne (1995) have pointed out the limitation of the

partial safety factor concept in nonlinear force calculations. If the forces

are computed in a nonlinear procedure, then an alternative definition has to

be chosen, such as

1 (7-121)

Rd = R( fcR ; f yR ; ftR ; f p 0,1 R ; f pR ) .

γR

Here, current work is being carried out. The reader should consult re-

cent publications, such as Cervenka (2007), Allaix et al. (2007), Holicky

(2007), and especially Pfeiffer and Quast (2003).

Several methods exist to develop partial safety factors for a certain mate-

rial or a certain type of structure. But in general, all procedures rely on a

statistical description of the material inherent uncertainty. Additionally,

historical partial safety factors remain valid if they have proven to provide

safe structures.

First, a historical procedure is introduced that permits the development of

partial safety factors for concrete only on the coefficient of variation, an

assumption about the type of probability distribution function, and the

class of the structure (Murzewski 1974) (Tables 7-19 and 7-20). Usually

the coefficient of variations for concrete depending on the production con-

ditions lies around 10% (Spaethe 1992, Östlund 1991).

Class of structure Coefficient of variation

Class of structure 0.05 0.10 0.15 0.20

Dam 1.26 1.58 1.95 2.43

Bridge, theatre, cultural buildings 1.23 1.49 1.82 2.16

Residential buildings, office buildings 1.20 1.40 1.65 1.91

Lager, bunker, frames 1.15 1.31 1.46 1.62

Secondary buildings 1.10 1.17 1.22 1.25

Coefficient of variation

Class of structure 0.05 0.10 0.15 0.20

Dam 1.24–1.30 1.46–1.85

Bridge, theatre, cultural buildings 1.21–1.26 1.41–1.67 1.60–2.47

Residential or office buildings 1.18–1.22 1.35–1.53 1.50–2.00 1.65–2.86

306 7 Safety Assessment

Coefficient of variation

Lager, bunker, frames 1.15–1.16 1.27–1.37 1.38–1.61 1.49–1.92

Secondary buildings 1.10–1.11 1.17–1.19 1.21–1.25 1.23–1.28

Joint Committee of Structural Safety (JCSS). In the background document

of the Eurocode provided by the JCSS (2004), an example for the devel-

opment of a partial safety factor is presented. The general description for

such a factor is

R = Θ⋅a⋅ x (7-122)

with Θ as uncertainty factor for the calculation model, a as geometrical

factor, and x as material strength. The resistance design value is then

Θ k ⋅ ak ⋅ x k (7-123)

Rd = .

γM

The three parameters might then be considered as independent random

variables with a lognormal distribution. Please note that this is not true for

many materials. The material partial safety factor can then be evaluated

according to

γ M = exp(α R ⋅ β ⋅ (Vx + 0.4 ⋅ Va + 0, 4 ⋅ VΘ ) − 1.64 ⋅ Vx ) . (7-124)

ues, including material partial safety factors based on testing of materials.

Such procedures have been intensively discussed by Reid (1999) and have

been applied for masonry (Curbach and Proske 2004). The procedures in-

clude some main assumptions; for example, the probability distribution

function of the load and the resistance. Examples are the Australian Stan-

dard Procedure for Statistical Proof Loading, Australian Standard Proce-

dure for Probabilistic Load Testing, or the Standard for Probabilistic Load

tests. Most of the procedures consider the variance of the material strength,

the variance of the load, some condition factors, the number of tests or a

correction factor, and the required safety index.

Val and Stewart (2002) describe a procedure for the evaluation of partial

safety factors for mainly the resistance of existing structures. The safety

factor is split into two parts, the partial safety factor for the material

strength and a factor for the consideration of additional uncertainties, fk,est is

the estimated characteristic material strength. The design value for the

strength is therefore

7.3 Semiprobabilistic Safety Concept 307

f k ,est (7-125)

fd = .

γ m ⋅ γη

Tables 7-21 and 7-22 show some examples taken from the publication

by Val and Stewart (2002). In Table 7-21, no a priori information was

available, whereas in Table 7-22, a priori information could be used. It be-

comes clear that additional information is for existing structures of major

importance for keeping safe and efficient structures. As already seen in the

JCSS (2004) example, the consideration of modelling uncertainty, in

which ever way, is a major part of many expansions of the traditional ways

for the developing partial safety factors. Here it becomes visible that the

clear rule for the application of statistics might sometimes be a drawback

for a realistic estimation of the safety of a structure.

Table 7-21. Example of material partial safety factors without a priori and with

six tested samples (Val and Stewart 2002)

Coefficient of variation of γm Coefficient of variation for the calibration factor

the samples 0.05 0.10 0.15

0.04 1.1 1.07 1.19 1.27

0.08 1.4 1.05 1.17 1.24

0.12 1.6 1.04 1.15 1.21

0.16 1.9 1.03 1.10 1.19

0.20 2.2 1.02 1.07 1.15

Table 7-22. Example of material partial safety factors with a priori and with six

tested samples (Val and Stewart 2002)

A priori coefficient of variation γm

0.05 1.36

0.10 1.25

0.15 1.21

cluded in the following material safety factor:

γ M = γ M 0 ⋅ ϕP ⋅ ϕD ⋅ ϕM . (7-126)

ration, and intensity of inspection.

Schneider (1999) has also investigated the development of partial safety

factors. Partial safety factors for existing reinforced concrete structures

308 7 Safety Assessment

have recently been discussed by Fischer and Schnell (2008) and Braml and

Keuser (2008).

Besides the simplified techniques, the partial safety factor can also be

computed using the results from full probabilistic computations (FORM).

Then, the so-called weighting factors αR of the random variables are used.

The partial safety factor for material strength γm is computed as

Rc (7-127)

γm = .

Rd

The safety factor can be computed when design resistance Rd and char-

acteristic resistance Rc are known. Since

Rd = μE − α R ⋅ β ⋅ σ R (7-128)

and for a normal distribution with known mean value and standard de-

viation, the characteristic value is given as

Rc = μE − 1.645 ⋅ σ R , (7-129)

hence the partial safety factor for the material strength can be computed.

As an example, Fig. 7-13 shows the distribution of the weighting factor

quadrates for historical arch bridges, either for road or for railway traffic

based on FORM computations. This is a common type of diagram since

the quadrates of the weighting factors have to sum up to one. The figure

clearly shows why in railway codes the partial safety factor for the traffic

load is often lower; for example, 1.3 compared to road traffic bridges with

1.5. Here it can be seen that the weighting factor is significantly lower for

railway traffic than for road traffic.

Table 7-23 shows the computation of partial safety factors for live and

dead load subjected to an adaptable goal safety index for arch bridges.

Table 7-24 lists system partial safety factors for stone arch bridges subject

to different masonry types. The investigation is based on own probabilistic

computations.

In general, the issue of partial safety factors for historical stone arch

bridges is still under discussion since they show a highly nonlinear behav-

iour and such structures can only be understood as a system and not on the

cross-section layer.

7.3 Semiprobabilistic Safety Concept 309

Fig. 7-13. Distribution of the square of the weighting factors for an arch bridge us-

ing a linear-elastic model for railway traffic (left) and road traffic (right)

Table 7-23. Partial safety factor for dead and live load considering the adaptation

of the safety index according to Schueremans and Van Gemert (2001)

Δ S + Δ R + Δ P + Δ I ≤ 1.5 Partial safety factor for Factor of combination

Dead load Live load

–0.25 1.42 1.56 0.69

0.00 1.35 1.50 0.70

0.25 1.28 1.45 0.71

0.50 1.22 1.39 0.72

0.75 1.16 1.34 0.73

1.00 1.10 1.29 0.74

1.25 1.05 1.25 0.75

1.50 1.00 1.20 0.76

Table 7-24. Own suggestions for partial safety factors for historical masonry arch

bridges subject to different masonry types in the arch

Masonry types

γ M = 2.1 γ M = 2.0 γ M = 1.9 γ M = 1.8

310 7 Safety Assessment

distribution. They relate the probability distribution property—for exam-

ple, strength—to a certain probability. A characteristic strength value fk of

structural materials of 5% of the overall population is very common. That

means that only 5% of the overall population will experience a lesser

strength, and 95% of the population will show a higher strength than this

5% fractile material strength. The terms percentile and quantile can often

also be found in literature instead of fractile. The chosen value of 5% is ar-

bitrary, but the general idea about characteristic values are the proofs in the

state of serviceability carried out without partial safety factor (one). Since

characteristic values and partial safety factors interact, the characteristic

value simply has to be chosen to provide the partial safety factor of one for

the limit state of serviceability. Therefore, the assumption of the 5% frac-

tile value can be found as a general requirement, for example, in the Euro-

code 1 (1994) or in the German DIN 1055-100, Sect. 6.4 (1999). As an ex-

ample, some material-related codes are listed here to show the wide

application of the 5% fractile value assumption:

• Reinforcement steel according to DIN 488 (90% confidence interval)

• Concrete compression strength according to DIN 1045, DIN 1048

• Masonry after testing according to DIN 1053 (75% confidence interval)

(Schubert 1995)

• Outer wall panelling according to DIN 18516 (75% confidence interval)

• Masonry according to DIN 18152 (1987) with 90% confidence interval

• Reinforcement steel according to ENV 100080 (90% confidence inter-

val)

• Wooden structures according to DIN V ENV 1995 (84,1%-confidence

interval, coefficient of variation greater or equal 0.1, more than 30 sam-

ples)

• Masonry-aerated concrete (95%-confidence interval)

• Artificial brick stones (Schubert 1995)

• Natural stones (90% confidence interval) (Schubert 1995)

To compute a 5% fractile value of a certain property, statistical informa-

tion is required. Using such statistical information, a probability distribu-

tion function can be estimated. There exists a wide variety of such prob-

ability functions as shown in Table 7-25. Many distribution functions can

be related to each other (Fig. 7-14). Other distribution functions describe

certain distribution families or certain conditions; for example, Piersons

differental equation and the Fleishmann system. An overview of distribution

7.3 Semiprobabilistic Safety Concept 311

families can be found in Plate (1993), Bobee and Ashkar (1991), and

Fischer (1999). Besides that for extreme value distributions, the work by

Castillo et al. (2008) is mentioned. For practical reasons, however, only a

limited number of distributions is considered for construction material: the

normal distribution, the lognormal distribution, and the Weibull distribu-

tion (Fischer 1999, Eurocode 1 1994, and GruSiBau 1981). The normal dis-

tribution has found wide application and can be easily explained by the

central limit state theorem. This theorem states that a sum of certain ran-

dom variables will yield to a normal distributed random variable if certain

conditions are fulfilled (Van der Werden 1957). Such conditions are, for

example, that no single random variable dominates the results. And indeed,

many material properties can be seen as the sum of certain other proper-

ties—for example, the strength of a natural stone can be seen as sum of the

strength of the single elements of the stone. Therefore, many publications

assume a normal distribution for the concrete compression strength (Rüsch

et al. 1969). A disadvantage of the normal distribution is possible negative

values. Therefore, instead of the normal distribution, the lognormal distri-

bution is often used if the average value of a material property is low and

experiences a high standard deviation, such as the tensile strength of con-

crete or masonry. The lognormal distribution does not feature negative

values and therefore no negative tensile strengths are then possible. The

lognormal distribution can also be related to the central limit state theorem,

if the data are logarithmic. However, this implies the multiplication of the

single input random variables—in other terms, the logarithmic distribution

fulfils the central limit state theorem for the case of multiplication. A fur-

ther often-used distribution for construction material properties is the

Weibull distribution (Weibull 1951). This distribution belongs to the group

of extreme value distributions and describes a chain interaction of single

elements. If the weakest part of the chain fails, then the entire chain fails,

which indeed fulfil’s the requirements of an extreme value distribution.

The properties of brittle materials, such as glass, can be described with this

distribution (Button et al. 1993, Güsgen et al. 1998). Of course, many ma-

terials do not comply with the chain rule for a serial system but show

rather a mixed parallel-serial system. Further considerations are then

needed. Some theoretical works dealing with this issue have been taken

out by Rackwitz and Hohenbichler (1981), Gollwitzer and Rackwitz

(1990), and Kadarpa et al. (1996) for brittle materials; Chudoba et al.

(2006) and Chudoba and Vorechovsky (2006) for glass yarns.

312 7 Safety Assessment

Name of distribution Name of distribution

1 χ-Distribution 29 Laplace distribution

2 General Pareto distribution 30 Logarithmic Pearson typ-3

distribution

3 Arcsin distribution 31 Logarithmic–logistic distribution

4 Beta distribution 32 Logistic distribution

5 Binomial distribution 33 Lognormal distribution

6 Birnbaum–Saunders distribution 34 Lorenz distribution

7 Breit–Wigner distribution 35 Maxwell distribution

8 Cauchy distribution 36 Neville distribution

9 Erlang distribution 37 Pareto distribution

10 Exponential distribution 38 Pearson, typ-3, gamma distribution

11 Extreme value distribution typ I max 39 Pearson, typ-3 distribution

12 Extreme value distribution typ I min 40 Poisson distribution

13 Extreme value distribution typ II max 41 Polya distribution

14 Extreme value distribution typ II min 42 Potential distribution

15 Extreme value distribution typ III max 43 Power normal distribution

16 Extreme value distribution typ III min 44 Rayleigh distribution

17 Fisher distribution 45 Uniform distribution

18 Fréchet distribution 46 Reverse Weibull distribution

19 F distribution 47 Rossi distribution

20 Gamma distribution (G-distribution) 48 Simpson or triangle distribution

21 Gauss order normal distribution 49 Sinus distribution

22 General extreme value distribution 50 Snedecor distribution

23 General Pareto distribution 51 Student-t-distribution

24 Geometric distribution 52 Tukey Lambda distribution

25 Gumbel distribution 53 Wakeby distribution

26 Hypergeometric distribution 54 Weibull distribution

27 Krickij–Menkel distribution 55 Wishart distribution

28 Landau distribution 56 Z-Distribution

7.3 Semiprobabilistic Safety Concept 313

However, theoretical considerations often do not prove the type of prob-

ability distribution. Therefore, usually the statistical data are investigated,

according to DIN 53804 (1981). The most robust and fast converting pa-

rameters of the random data are general tendency estimators such as the ar-

ithmetical mean, harmonically mean, geometrical mean, generalized mean,

quadratic mean and median (50% fractile), and mode (most probable

value). However, more interesting from a statistical point of view, are de-

viation measures or uncertainty measures such as variance, standard devia-

tion (same unit as mean value), coefficient of variation, span, modified in-

terquartile range, and mean deviation. Besides such measures, the

skewness and the kurtosis are also of interest (Fig. 7-15).

Fig. 7-15. Meaning of the statistical parameter arithmetic mean (Xm), standard de-

viation (σ), skewness (S), and kurtosis (K)

314 7 Safety Assessment

For historical masonry, one should keep in mind that such statistical pa-

rameters may be insufficient due to corrupted data (Fig. 7-16). Statistical

data from historical masonry with natural stone may feature outliers, cen-

sored data, and multimodal data. Outliers are samples that do not belong to

the population. However, since we do not know the population, it is diffi-

cult to identify outliers. Here, Dixons test, Barnett–Lewis test, Chauvenets

criteria, Grupps test, the David-Hartley-Pearson test and other criteria may

be used for outlier identification (McBean and Rovers 1998, Fischer 1999,

and Bartsch 1991). Censored data describe a condition in which access to the

original population data is filtered by a process. For example, investigating

the strength of historical mortar may yield to censored data since the drill-

ing process using high-pressure cooling water may destroy mortar inside

the masonry. Furthermore, sawing the test specimen may further destroy

material. Therefore, the compression test results of the mortar may indicate

high average compression strength, however the tests do not consider the

failure of all weak material in the preparation process. Therefore, the data

are censored and have to be corrected. Then, for example, Cohens and

Aitchison’s method can be applied for data correction (McBean and Rov-

ers 1998). Finally, it may be the case that the test results do not come from

one population. For example, different stone types may be used for maso-

nry or the natural stone material may come from different stone quarries.

Then the compression strength can show multimodal behaviour. Using op-

timization methods, it is possible to disintegrate different distributions

from such multimodal data (Proske 2003).

function usually has to be chosen. Several statistical techniques exist to in-

vestigate statistical data and recommend a distribution type:

• Relating the coefficient of variance and the type of distribution

• Relating skewness and kurtosis and the type of distribution (Fig. 7-17)

• The minimum sum square error based on histograms

• The χ test and nω test

2 2

• The Shapiro-Wilk or Shapiro-Francia test

7.3 Semiprobabilistic Safety Concept 315

• Probability plots

• Quantile-correlation values

Such statistical tests are often called goodness-of-fit tests. However, as

described, statistical testing is only one part of the investigation. The un-

derstanding of the material extraction, the testing techniques, and the ma-

terial itself is compelling to interpret the results of material testing and sta-

tistical investigations.

Fig. 7-17. Relation between skewness and kurtosis and the type of distribution

(Plate 1993) D. = distribution

random variables is often of great interest in identifying deterministic formu-

las. However, the identification of different correlation coefficients is rather

laborious due to the required high sample sizes. Figure 7-18 shows that

measuring a correlation coefficient of 0.5 for ten samples shows a value for

the population between –0.5 and 0.8. Furthermore, Pearson’s coefficient of

correlation may not be adequate, other correlation coefficients such as from

Spearman, Kendall, and Hoeffding may be helpful. Additionally, nonlinear

regression using the Levenberg-Marquardt method can be considered.

Besides such limitations, the next sections will discuss the computation of

characteristic 5% fractile values for certain probability distribution functions.

316 7 Safety Assessment

(Steel and Torrie 1991)

The 5% fractile value fk of a normal distributed material strength can be

computed as (Storm 1988)

fk = fm − k ⋅ σ , (7-130)

where fm is the mean value and σ is the standard deviation. If the mean

value and the standard deviation are known, then the k-factor correspond-

ing to the 5% fractile value amounts to 1.645. However, usually only some

suggestions for the mean value and the standard deviation are known due

to a limited number of samples. Although the mean value converges very

fast with a low number of samples, the problem remains for the evaluation

of the standard deviation. The uncertainty in the empirical statistical pa-

rameters is usually considered in the choice of the k-factor.

Assuming, for example, a normal distributed property with 15 samples

and a confidence interval of 95%, the k-factor becomes 1.76 based on a

student-t distribution (Table 7-26). For very high sample numbers, the k-

value converges to 1.645 again.

7.3 Semiprobabilistic Safety Concept 317

sample number – 1)

freedom 5% 2.5% freedom 5% 2.5%

1 6.314 12.706 20 1.725 2.086

2 2.920 4.303 21 1.721 2.080

3 2.353 3.182 22 1.717 2.074

4 2.132 2.776 23 1.714 2.069

5 2.015 2.571 24 1.711 2.064

6 1.943 2.447 25 1.708 2.060

7 1.895 2.365 26 1.706 2.056

8 1.860 2.306 27 1.703 2.052

9 1.833 2.262 28 1.701 2.048

10 1.812 2.228 29 1.699 2.045

11 1.796 2.201 30 1.697 2.042

12 1.782 2.179 40 1.684 2.021

13 1.771 2.160 60 1.671 2.000

14 1.761 2.145 80 1.664 1.990

15 1.753 2.131 100 1.660 1.984

16 1.746 2.120 200 1.653 1.972

17 1.740 2.110 500 1.648 1.965

18 1.734 2.101 1000 1.646 1.962

cording to the Eurocode 1, 15 samples would yield a k-factor of 1.84.

A simple example should illustrate the application of a normal distribu-

tion. About 500 compression test results of Posta sandstone are available

for an investigation. The mean value of the compression strength is 58.05

MPa and the standard deviation is given with 10.32 MPa. Assuming a

normal distribution for 500 samples, the k-value becomes 1.648. The char-

acteristic compression strength then is computed as

f st ,k = 58.05 MPa − 1.648 ⋅ 10.32 MPa = 41.04 MPa . (7-131)

sion strength of Posta sandstone. Since the sample size is unknown, the k-

value is kept constant with 1.645.

Grunert et al. (1998) f st ,k = f m − 1.645 ⋅ σ = 31.6 − 1.645 ⋅ 6.8 = 20.41 MPa

Peschel (1984) f st ,k = f m − 1.645 ⋅ σ = 41.6 − 1.645 ⋅ 11.6 = 22.52 MPa

318 7 Safety Assessment

As mentioned above, the lognormal distribution can be reasoned based on

a product formulation of the central limit theorem. The computation of the

5% fractile value is analogous to the normal distribution, except that the

data have to be transformed by the logarithm:

L ' = y − k ⋅ s* = ln( L ) (7-132)

n

yi (7-133)

y =∑ with yi = ln( xi )

i =1 n

n (7-134)

∑(y − y)

2

i

s* = i =1

.

n −1

The following example should show the application. Four compression

strength tests from natural stone masonry have been carried out (Table 7-27).

Sample Compression strength in MPa Logarithm

1 6.70 1.90

2 6.20 1.82

3 5.70 1.74

4 6.60 1.89

Mean value 6.30 1.84

Standard deviation 0.39 0.06

puted as follows (for 2.353 see Table 7-26):

f mw ,k = exp(1.84 − 2.353 ⋅ 0.06) = 5.47 MPa . (7-135)

comes

f mw ,k = 6.30 MPa − 2.353 ⋅ 0.39 MPa = 5.37 MPa . (7-136)

For the Weibull distribution, the 5% fractile value can be estimated with

7.3 Semiprobabilistic Safety Concept 319

1 (7-137)

⎛ 1 ⎞k

fk = ⎜ − ln(1 − q ) ⎟ .

⎝ λ ⎠

The value q describes the probability that is chosen in our case as 0.05

for the 5% fractile value. The factors λ and k are parameters of the Weibull

distribution, which can be computed based on statistical data. The mean

value and the standard deviation can be computed as

⎛ 1⎞ (7-138)

f m = f0 + λ −1/ k ⋅ Γ ⎜ 1 + ⎟ ,

⎝ k⎠

2 (7-139)

⎛ 2⎞ ⎛ 1⎞

σ =λ −1/ k

Γ ⎜1 + ⎟ − Γ ⎜1 + ⎟ .

⎝ k⎠ ⎝ k⎠

Since the mean value and the standard deviation of the test data can

usually be easily computed, these formulas are useful in computing the

factors λ and k.

Again, an example shows the application. Consider the test results of the

compression strength of natural stone masonry shown in Table 7-27. Using

the mean value of 6.30 MPa and the standard deviation of 0.39 MPa, the k

value can be computed with 20.01 and the λ becomes 5.84 × 10 . The 5%

–17

1 (7-140)

⎛ 1 ⎞ 20.01

fk = ⎜ − −17

ln(0.95) ⎟ = 5.58 MPa .

⎝ 5.84 ⋅10 ⎠

The Leicester method estimates the 5% fractile value without the selection

of a probability function (Hunt and Bryant 1996). The 5% fractile value is

computed as

⎛ 2.7 ⋅ v ⎞ (7-141)

fk = A ⋅ ⎜1 − ⎟

⎝ n ⎠

with

n as number of samples preferred higher than 30

v as coefficient of variation preferred smaller than 0.5

A as empirical 5% fractile of the data. This value is often computed by

linear interpolation.

320 7 Safety Assessment

stone compression strength tests, the empirical 5% fractile value of the

sample data is estimated with 42.3 MPa. Such estimations can, for example,

be done with many spreadsheet programs such as Excel. In Excel, the

“rang and quantile” function can be used. The coefficient of variation is

0.177. Then, the corrected 5% fractile value is given by

⎛ 2.7 ⋅ 0.177 ⎞ (7-142)

f k = 42.3 MPa ⋅ ⎜ 1 − ⎟ = 42.26 MPa .

⎝ 500 ⎠

A further technique is the Öfverbeck Power Limit (Hunt and Bryant 1996).

Using the constant ε, subject to the number of samples and the number of

used samples q, the 5% fractile value is given as

q −1 (7-143)

fk = x1q−ε ∏ xiε /( q −1) .

i =1

Table 7-28. Relevant constants q and ε subject to the overall sample size

Sample size n Number of used samples q Öfverbeck constant ε

5 2 5.93

6 2 5.35

7 2 4.85

8 2 4.42

9 2 4.03

10 3 3.31

11 3 3.12

12 3 2.96

13 3 2.80

14 3 2.66

15 3 2.53

20 4 2.22

30 5 1.80

40 6 1.58

50 7 1.44

Only four samples are available. Using Table 7-28, q is 2 and 6.00 is chosen

7.3 Semiprobabilistic Safety Concept 321

for ε by extrapolation. The samples are given by 5.7, 6.2, 6.6, and 6.7

MPa. The 5% fractile value is then

q −1 2 −1 (7-144)

fk = x1q−ε ∏ xiε /( q −1) = x12−6.00 ∏ xi

i =1 i =1

6

5.7

= 6.21−6 ⋅ 5.76 /(2 −1) = = 3.74 MPa .

6.25

Jaeger and Bakht (1990) present a technique to estimate the 5% fractile

based on a combination of different probability functions. They call this

artificial function a log arc sinh normal polynomial distribution. Assuming

a quadratic polynomial in the distribution, and the consideration of only

the lowest, average and highest test value, the fractile value can be com-

puted with the following steps:

1. Sort the data.

2. Compute the mean value. If the sample number is uneven, then use

the median. If the sample number (n=2 × k) is even, use the follow-

ing formula:

( f k + f k +1 ) (7-145)

fm = .

2

3. Transform the data according to

fi 2 − fm2 (7-146)

yi = .

2 ⋅ fi ⋅ f m

4. Chose the characteristic value—in our case, the 5% fractile value.

Then chose the representing k-value of a normal distribution, here k

= z* = –1,645.

5. Chose a z1 according to Table 7-29.

Sample number n z1

10 1.34

11 1.38 As approximation, the following equation

12 1.43 can be recommended:

13 1.47 z1 = −0.8004 − 0.0649 ⋅ n + 0.0011 ⋅ n2

14 1.50

15 1.53

16 1.56

322 7 Safety Assessment

Sample number n z1

17 1.59

18 1.62

19 1.65

20 1.67

2 (7-147)

⎛ y − y ⎞⎛ z ⎞ ⎛ y + y ⎞⎛ z ⎞

* *

y* = ⎜ 1 n ⎟ ⎜ ⎟ + ⎜ 1 n ⎟⎜ ⎟ .

⎝ 2 ⎠ ⎝ z1 ⎠ ⎝ 2 ⎠ ⎝ z1 ⎠

7. Compute the 5% fractile value for the original data:

f k = f m ( y* + 1 + ( y* )2 . (7-148)

and Bakht (1990); however, they consider this one as the numerically most

robust one. The other techniques consider other data points, such as the

smallest, the second smallest, and the mean value, or other orders of the

polynomial, and yield slightly different 5% fractile values.

Again, an example illustrates the approach. Fifteen bending test results

of granite stones are used. The test results are then related to one specimen

height and listed in Table 7-30.

No. Bending tensile strength in MPa Percent yi

1 14.710 100.00 0.35106

2 12.802 92.80 0.20674

3 12.700 85.70 0.19858

4 12.506 78.50 0.18291

5 11.150 71.40 0.06719

6 10.829 64.20 0.03793

7 10.822 57.10 0.03729

8 10.426 50.00 0.00000

9 10.208 42.80 –0.02113

10 10.002 35.70 –0.04153

11 9.919 28.50 –0.04987

12 9.664 21.40 –0.07597

13 9.412 14.20 –0.10250

14 6.281 7.10 –0.52875

15 4.233 0.00 –1.02851

References 323

transformed data (Step 3) are already included in Table 7-30. Furthermore,

the 5% fractile value should be computed, therefore z* = –1.645 and z1=

1.53. Then the transformed 5% fractile value is given as

2 (7-149)

y =⎜

*

⎟⎜ ⎟+⎜ ⎟⎜ ⎟

⎝ 2 ⎠ ⎝ −1.53 ⎠ ⎝ 2 ⎠⎝ −1.53 ⎠

y* = −1.133

and for the original data

( )

f k = 10.426 MPa ⋅ −1.133 + 1 + ( −1.133)2 = 3.94 MPa . (7-150)

Mehdianpour (2006) has used diagrams of binomal distribution to estimate

characteristic compression strength values. Details can be found in the

original reference.

All mentioned techniques consider the uncertainty of material and load-

ing data in a stochastic way. However, such a consideration does not nec-

essarily end in the preparation of fractile values and partial safety factors.

Increasingly, full probabilistic computations of structures and also of his-

torical stone arch bridges can be found in the literature. Therefore, Chapter 8

discusses the results of those numerical investigations.

References

Allaix DL, Carbone VI & Mancini G (2007) Global Safety Factors for Reinforced

Concrete Beams. Proceedings of the 5th International Probabilistic Workshop,

L Taerwe & D Proske (Eds), Ghent, Belgium

Bamboo (2005) The Hong Kong Polytechnic University and International Network

for Bamboo and Rattan: Bamboo Scaffolds in Building Construction – Design

Guide. www.inbar.int/publication/txt/INBAR_Technical_Report_No23.doc

Bartsch HJ (1991) Mathematische Formeln. 23. Auflage, Fachbuchverlag Leipzig

Bayer (2008) Efficient Modelling and Simulation of Random Fields. Presentation

at the 6th International Probabilistic Workshop, Darmstadt

Bayer V & Ross D (2008) Efficient Modelling and Simulation of Random Fields.

Proceedings of the 6th International Probabilistic Workshop, CA Graubner, H

Schmidt & D Proske (Eds), Darmstadt, pp 3363–380

Bergman LA, Shinozuka M, Bucher CG, Sobczyk K, Dasgupta G, Spanos PD,

Deodatis G, Spencer Jr. BF, Ghanem RG, Sutoh A, Grigoriu M, Takada T,

Hoshiya M, Wedig WV, Johnson EA, Wojtkiewicz SF, Naess A, Yoshida I,

324 7 Safety Assessment

report on computational stochastic mechanics. Probabilistic Engineering

Mechanics, 12 (4), pp 197-321

Bergmeister K (2003) Kohlenstofffasern im Konstruktiven Ingenieurbau. Verlag

Ernst und Sohn, Berlin

Bobee B & Ashkar F (1991) The Gamma Family and Derived Distributions

Applied in Hydrology. Water resources publications, Littleton

Bothe E, Henning J, Curbach M, Bösche T & Proske D (2004) Nichtlineare

Berechnung alter Bogenbrücken auf der Grundlage neuer Vorschriften. Beton-

und Stahlbetonbau 99, Heft 4, pp 289–294

Box GEP & Draper NR (1987) Empirical Model-Building and Response Surfaces.

John Wiley & Sohn

Braml T & Keuser M (2008) Structural Reliability Assessment for Existing Bridges

with In Situ Gained Structural Data Only – Proceeding and Assessment of

Damaged Bridges. Proceedings of the 6th International Probabilistic Workshop,

CA Graubner, H Schmidt & D Proske (Eds), Darmstadt, pp 149–165

Breitung K (1984) Asymptotic approximations for multinormal integrals. Journal

of the Engineering Mechanics Division, 110 (3), pp 357–366

Breitung K (2002) SORM: Some common misunderstandings and errors. In:

Application of Statistics and Probability in Civil Engineering, A Der Kiureghian,

S Madanat & JM Pestana (Eds), Millpress, Rotterdam 2003, pp 35–40

Bucher C & Bourgund U (1990) A fast and efficient response surface approach for

structural reliability problems. Structural Safety, 7, pp 57–66

Bucher C (1988) Adaptive sampling – An iterative fast Monte Carlo procedure.

Structural Safety, 5, pp 119–126

Bucher C, Hintze D & Roos D (2000) Advanced analysis of structural reliability

using commercial FE-codes. European Congress on Computational Methods

in Applied Sciences and Engineering (ECCOMAS), CIMNE, Barcelona, 11–14

September 2000

Busch P & Zumpe G (1995) Tragfähigkeit, Tragsicherheit und Tragreserven von

Bogenbrücken. 5. Dresdner Brückenbausymposium, 16.3.1995, Fakultät

Bauingenieurwesen, Technische Universität Dresden, pp 153–169

Button D et al. (1993) Glass in Building – A Guide to Modern Architectural Glass

Performance. Pilkington Glass Ltd. Flachglas AG, Pilkington

BÜV (2001) Empfehlung für die Bemessung und Konstruktion von Glas im

Bauwesen. veröffentlicht in: Der Prüfingenieur 18 April 2001, pp 55–69

Cai GQ & Elishakoff I (1994) Refined second-order reliability analysis. Structural

Safety, 14, pp 267–276

Casas JR, Prato CA, Huerta S, Soaje PJ & Gerbaudo CF (2001) Probabilistic

assessment of roadway and railway viaducts. Safety, Risk, Reliability –

Trends in Engineering, Malta 2001, pp 1001–1008

Castillo E, Castillo C & Mínguez R (2008) Use of extreme value theory in

engineering design. Safety, Reliability and Risk Analysis: Theory, Methods

and Applications, S Martorell et al. (Eds), Taylor & Francis Group, London,

pp 2473–2488

References 325

standard – codes of practice for structures. Volume I: Common unified rules

for different types of construction and material (3rd Draft, Master Copy),

Bulletin d’information 116 E, Paris, November 1976

Cervenka V (2007) Global Safety Format for Nonlinear Calculation of Reinforced

Concrete. Proceedings of the 5th International Probabilistic Workshop,

L Taerwe & D Proske (Eds), Ghent, Belgium

Chudoba R & Vořechovský M (2006) Stochastic modeling of multi-filament

yarns: II. Random properties over the length and size effect, International

Journal of Solids and Structures, 43 (3-4), pp 435–458.

Chudoba R, Vořechovský M & Konrada M (2006) Stochastic modeling of multi-

filament yarns. I. Random properties within the cross-section and size effect,

International Journal of Solids and Structures, 43 (3-4), pp 413–434

COST-345 (2004) European Commission Directorate General Transport and

Energy: COST 345 – Procedures Required for the Assessment of Highway

Structures: Numerical Techniques for Safety and Serviceability Assessment –

Report of Working Groups 4 and 5

Curbach M & Proske D (2004) Zur Ermittlung von Teilsicherheitsfaktoren für

Natursteinmaterial. 12. November 2004, 2. Dresdner Probabilistik Symposium.

Technische Universität Dresden. pp 99–128

Curbach M, Michler H & Proske D (2002) Application of Quasi-Random

Numbers in Monte Carlo Simulation. 1st International ASRANet Colloguium.

8–10th July 2002, Glasgow, auf CD

DAfSt – Deutscher Ausschuß für Stahlbeton (1996) Richtlinie für Betonbau beim

Umgang mit wassergefährdenden Stoffen, Beuth-Verlag: Berlin

Der Kiureghian A & Ke J-B (1988) The stochastic finite element method in

structural reliability. Probabilistic Engineering Mechanics, 3 (2), pp 105–112

Der Kiureghian A, Haukaas T & Fujimura K (2006) Structural reliability software

at the University of California, Berkeley. Structural Safety 28, pp 44–67

Diamantidis D, Holicky M & Jung K (2007) Assessment of existing structures –

On the applicability of the JCSS recommendations. In: Aspects of Structural

Reliability – In Honor of R. Rackwitz, MH Faber, T Vrouwenvelder, K Zilch

(Eds), Herbert Utz Ver-lag, München

DIN 1045-1 (2001) Tragwerke aus Beton, Stahlbeton und Spannbeton, Teil 1:

Bemessung und Konstruktion, Juli 2001

DIN 1053-100 (2004) Mauerwerk – Berechnung auf der Grundlage des

semiprobabilistischen Sicherheitskonzeptes. NABau im DIN e.V.: Berlin

DIN 1055-100 (1999): Einwirkungen auf Tragwerke, Teil 100: Grundlagen der

Tragwerksplanung, Sicherheitskonzept und Bemessungsregeln, Juli 1999

DIN 18 516 (1999) Außenwandbekleidungen

DIN 53 804 (1981) Teil 1: Statistische Auswertungen – Meßbare (kontinuierliche)

Merkmale. September 1981

Drezner Z (1992) Computation of the multivariate normal integral. ACM

Transaction on Mathematical Software, 18 (4), pp 470–480

Eibl J & Schmidt-Hurtienne B (1995) Grundlagen für ein neues Sicherheitskonzept,

Bautechnik, 72, Nr. 8, pp 501–506

326 7 Safety Assessment

87, Heft 6, pp 137–139

EN 1994 (2006) Eurocode 4: Design of composite steel and concrete structures

EN 1995 (2008) Eurocode 5: Design of timber structures

EN 1996 (2006) Eurocode 6: Design of masonry structures

EN 1999 (2007) Eurocode 9: Design of aluminium structures

Epstein S, Rauzyb A & Reinhart F M (2008) The Open PSA Initiative for Next

Generation Probabilistic Safety Assessment. Proceedings of the 6th

International Probabilistic Workshop, CA Graubner, H Schmidt & D Proske

(Eds), Darmstadt

Estes AC & Frangopol DM (1998) RELSYS: A computer program for structural

system reliability. Structural Engineering and Mechanics, 6 (8), pp 901–919

Eurocode 1 (1994) (ENV 1991 –1): Basis of Design and Action on Structures, Part

1: Basis of Design. CEN/CS, August 1994

Fießler B, Hawranek R & Rackwitz R (1976) Numerische Methoden für

probabilistische Bemessungsverfahren und Sicherheitsnachweise. Technische

Universität München, Berichte zur Zuverlässigkeitstheorie, Heft 14

Fischer A & Schnell J (2008) Determination of partial safety factors for existing

structures. Proceedings of the 6th International Probabilistic Workshop, CA

Graubner, H Schmidt & D Proske (Eds), Darmstadt, pp 133–147

Fischer L (1999) Sicherheitskonzept für neue Normen – ENV und DIN-neu,

Grundlagen und Hintergrundinformationen. Teil 3: Statistische Auswertung

von Stichproben im eindimensionalen Fall. Bautechnik 76 (1999), Heft 2,

pp 167–179, Heft 3, pp 236–251, Heft 4, pp 328–338

Flederer H (2001) Beitrag zur Berechnung durchlaufender Stahlverbundträger im

Gebrauchszustand unter Berücksichtigung streuender Eingangsgrößen.

Dissertation, Technische Universität Dresden, Lehrstuhl für Stahlbau

Franz G, Hampe E & Schäfer K (1991) Konstruktionslehre des Stahlbetons. Band

II: Tragwerke, Zweite Auflage, Springer Verlag, Berlin

Freudenthal AM (1947) Safety of Structures, Transactions ASCE, V. 112, 1947,

pp 127–180

Genz A (1992) Numerical computation of multivariate normal probabilities,

Journal of Computational and Graphical Statistics, 1, pp 141–149

Ghanem G & Spanos PD (1991) Stochastic Finite Element: A Spectral Approach,

publisher, Springer, New York

Gollwitzer S & Rackwitz R (1990) On the reliability of Daniels systems.

Structural Safety, 7, pp 229–243

Gollwitzer S, Kirchgäßner B, Fischer R & Rackwitz R (2006) PERMAS-

RA/STRUEL system of programs for probabilistic reliability analysis.

Structural Safety 28, pp 108–129

Grunert B, Grunert S & Grieger C (1998) Die historischen Baustoffe der

Marienbrücke zu Dresden. Wissenschaftliche Zeitschrift der Technischen

Universität Dresden, 47, Heft 5/6, pp 11–20

Grunert S (1982) Der Sandstein der Sächsischen Schweiz als Naturressource,

seine Eigenschaften, seine Gewinnung und Verwendung in Vergangenheit

und Gegenwart. Dissertation, Technische Universität Dresden

References 327

von Sicherheitsanforderungen für bauliche Anlagen. Beuth Verlag

Güsgen J, Sedlacek G & Blank K (1998) Mechanische Grundlagen der

Bemessung tragender Bauteile aus Glas, Stahlbau 67, pp 281–292

Hasofer AM & Lind MC (1974) Exact and invariant second moment code format,

Journal of Engineering Mechanics Division, 100, pp 111–121

Holicky M (2007) Probabilistic Approach to the Global Resistance Factor for

Reinforced Concrete Members. Proceedings of the 5th International

Probabilistic Workshop, L Taerwe & D Proske (Eds), Ghent, Belgium

Hunt RD & Bryant AH (1996) Statistical Implications of Methods of Finding

Characteristic Strengths. Journal of Structural Engineering, 122 (2), pp.

202–209

Ibrahim Y (1991) Observations on applications of importance sampling in

structural reliability analysis, Structural Safety, 9, pp 269–281

ISO/CD 13822 (1999) Bases for design of structures – Assessment of existing

structures. International Organisation for Standardization, Geneva

Jaeger LG & Bakht B (1990) Lower Fractiles of Material Strength from limited

test data. Structural Safety, 7, pp 67–75

Jäger W (2005) Zur Einführung der DIN 1053-100, Mauerwerk – Berechnung auf

der Grundlage des semiprobabilistischen Sicherheitskonzepts. Mauerwerk 9,

1, pp 21–23

JCSS (2004) – Joint Committee of Structural Safety. Probabilistic Modelcode.

www.jcss.ethz.ch

Kandarpa S, Kirkner DJ & Spencer BF (1996) Stochastic damage model for

Brittle materials subjected to monotonic loading. Journal of Engineering

Mechanics, 122 (8), Aug. 1996, pp 788–795

Klingmüller O & Bourgund U (1992) Sicherheit und Risiko im Konstruktiven

Ingenieurbau. Braunschweig/Wiesbaden: Vieweg

Köylüoglu HU & Nielsen SRK (1994) New approximations for SORM integrals.

Structural Safety, 13, pp 237–246

Lemaire M & Pendola M (2006) PHIMECA-Soft. Structural Safety, 28, pp 130–149

Lin H-Z & Khalessi MR (2006) General outlook of UNIPASS V5.0: A general

purpose probabilistic software system. Structural Safety, 28, pp 196–216

Liu P & Der Kiureghian A (1986) Multivariate distribution models with

prescribed marginals and covariances. Probabilistic Engineering Mechanics,

1, pp 105–112

Low BK & Teh CI (2000) Probabilistic Analysis of Pile Deflection under Lateral

Loads. In: Application of Statistics and Probability ICASP 8, RE Melchers &

MG Stewart (Eds), Volume 1, Balkema, Rotterdam, pp 407–414

Macke M (2000) Variance reduction in Monte Carlo Simulation of dynamic

systems. Applications of Statistics and Probability (ICASP 8). RE Melchers &

MG Stewart (Eds), Balkema, Rotterdam, Band II, pp 797–804

Maes MA, Breitung K & Dupuis DJ (1993) Asymptotic importance sampling.

Structural Safety, 12, pp 167–186

Mann W (1987) Überlegungen zur Sicherheit im Mauerwerksbau. Mauerwerks-

Kalender. Verlag Ernst & Sohn, Berlin, pp 1–5

328 7 Safety Assessment

Mann W (1999) Mauerwerk in Europa. Der Prüfingenieur. 14. April 1999, VPI,

pp 53–59

Mayer M (1926) Die Sicherheit der Bauwerke und ihre Berechnung nach

Grenzkräften anstatt nach zulässigen Spannungen. Verlag von Julius Springer,

Berlin

McBean EA & Rovers FA (1998) Statistical procedures for analysis of environ-

mental monitoring data & risk assessment. Prentice Hall PTR Environmental

Management & Engineering Series, Volume 3, Prentice Hall, Inc., Upper

Saddle River

Mehdianpour M (2006) Tragfähigkeitsbewertung aus Versuchen – Probenanzahl

versus Aussagesicherheit. Proceedings of the 4th International Probabilistic

Symposium, D Proske, M Mehdianpour & L Gucma (Eds), Berlin, 2006

Melchers RE & Faber MH (2001) Aspects of Safety in Design and Assessment of

Deteriorating Structures, Proceedings to the International IABSE Conference

on Safety, Risk and Reliability - Trends in Engineering, March 21-23, Malta

2001, pp. 161-166

Melchers RE (1999) Structural Reliability Analysis and Prediction. 2nd Edition,

John Wiley & Sons

Mori Y & Ellingwood BR (1993) Time-dependent system reliability analysis by

adaptive importance sampling. Structural Safety, 12, pp 59–73

Most T (2008) An adaptive response surface approach for reliability analyses of

discontinuous limit state functions. Proceedings of the 6th International

Probabilistic Workshop, CA Graubner, H Schmidt & D Proske (Eds), Darmstadt,

pp 381–396

Murzewski J (1974) Sicherheit der Baukonstruktionen. VEB Verlag für Bauwesen,

Berlin, DDR

Norton RL (1996) Machine Design – An Integrated Approach. Prentice-Hall,

New York

NR (1992) Numerical Recipes in FORTRAN 77: The Art of Scientific Computing.

Cambridge University Press

Onken P, vom Berg W & Neubauer U (2002) Verstärkung der West Gate Bridge,

Melbourne, Beton- und Stahlbetonbau, Heft 2, 97 Jahrgang, pp 94–104

Östlund L (1991) An Estimation of gamma-Values. An application of a

probabilistic method. In: Reliability of Concrete Structures. Final report of

Permanent Commission I. CEB Bulletin d’ Information. Heft 202, pp 37–51

Peschel A (1984) Natursteine der DDR – Kompendium petrographischer, petro-

chemischer, petrophysikalischer und gesteinstechnischer Eigenschaften.

Beilage zur Dissertation (B) an der Bergakademie Freiberg

Petschacher M (1994) VaP a Tool for Practicing Engineers. Proceedings of

ICOSSAR’93, GI Schueller et al. (Ed), 6th International Conference on

Structural Safety and Reliability in Innsbruck, Balkema, pp 1817–1823

Pfeiffer U & Quast U (2003) Nichtlineares Berechnen stabförmiger Bauteile.

Beton- und Stahlbetonbau 98, Volume 9, pp 529–538

Plate EJ (1993) Statistik und angewandte Wahrscheinlichkeitsrechnung für

Bauingenieure, Ernst & Sohn Verlag, Berlin

Polidori DC, Beck JL & Papadimitriou C (1999) New approximations for reli-

ability integrals. Journal of Engineering Mechanics, April 1999, pp 466–475

References 329

Proske D (2003) Ein Beitrag zur Risikobeurteilung von alten Brücken unter

Schiffsanprall. Ph.D. Thesis, Technische Universität Dresden

Proske D (2008a) Definition of Safety and the Existence of “Optimal Safety”.

Proceedings of the ESREL 2008 conference, Valencia

Proske D (2008b) Catalogue of risks – Natural, technical, health and social risks.

Springer, Berlin – Heidelberg

Proske D, Michler H & Curbach M (2002) Application of quasi-random numbers

in Monte Carlo Simulation. ASRANET, Glasgow, on CD

Pukl R, Novak D, Vorechovsky M & Bergmeister K (2006) Uncertainties of

Material Properties in Nonlinear Computer Simulation. Proceedings of the 4th

International Probabilistic Symposium, D Proske, M Mehdianpour und

L Gucma (Eds), Berlin, pp 127–138

Pukl R, Novák D, Vořechovský M & Bergmeister K (2006) Uncertainties of

material properties in nonlinear computer simulation. D Proske, M

Mehdianpour & L Gucma (Eds), Proceedings of the 4th International

Probabilistic Symposium, 12th–13th October 2006, Berlin

Rachwitz R & Hohenbichler M (1981) An Order Statistics Approach to Parallel

Structural Systems, Berichte zur Zuverlässigkeitstheorie der Bauwerke, SFB

96, Technische Universität München, Heft 58

Rackwitz R & Hohenbichler M (1981) An Order Statistics Approach to Parallel

Structural Systems, Berichte zur Zuverlässigkeitstheorie der Bauwerke, SFB

96, Technische Universität München, Heft 58

Rajashekhar MR & Ellingwood BR (1993) A new look at the response surface

approach for reliability analysis, Structural Safety, 12, pp 205–220

Reh S, Beley J-D, Mukherjee S & Khor EH (2006) Probabilistic finite element

analysis using ANSYS. Structural Safety, 28, pp 17–43

Reid SG (1999) Load–testing for design of structures with Weibull–distributed

strengths. Application of Statistics and Probability (ICASP 8), Sydney, 1999,

Band 2, pp 705–712

Roos D & Bayer V (2008) An efficient robust design optimization approach using

advance surrogate models. Proceedings of the 6th International Probabilistic

Workshop, CA Graubner, H Schmidt & D Proske (Eds), Darmstadt

Ross D & Bucher C (2003) Adaptive response surfaces for structural reliability of

nonlinear finite element structures. NAFEMS Seminar: Use of Stochastics in

FEM Analyses. May 7–8 2003 Wiesbaden

Rüsch R, Sell R & Rackwitz R (1969) Statistische Analyse der Betonfestigkeit.

Deutscher Ausschuß für Stahlbeton, Heft 206, Berlin

Schervish MJ (1984) Multivariate normal probabilities with error bound. Applied

Statistics, Royal Statistical Society, 33 (1), pp 81–94

Schlegel R & Will J (2007) Sensitivitätsanalyse und Parameteridentifikation von

bestehenden Mauerwerkstrukturen. Mauerwerk 11, Heft 6, pp 349–355

Schleicher F (Ed) (1949) Taschenbuch für Bauingenieure, Springer-Verlag

Schneider J (1999) Zur Dominanz der Lastannahmen im Sicherheitsnachweis.

Festschrift zum 60. Geburtstag von Prof. Dr. Edoardo Anderheggen. Institut

für Baustatik und Konstruktion, Eidgenössische Technische Hochschule

Zürich, pp 31–36

330 7 Safety Assessment

aus künstlichen Steinen und Naturstein. Mauerwerkskalender 1995, pp 687–701

Schueller GI & Pradlwarter HJ (2006) Computational stochastic structural

analysis (COSSAN) – a software tool. Structural Safety, 28, pp 68–82

Schueremans L & Van Gemert D (2001) Assessment of existing masonry

structures using probabilistic methods - state of the art and new approaches.

STRUMAS 2001, 18.–20. April 2001, Fifth International Symposium on

Computer Methods in Structural Masonry

Schweckendiek T & Courage W (2006) Structural reliability of sheet pile walls

using finite element analysis. Proceedings of the 4th International Probabil-

istic Symposium, D Proske, M Mehdianpour & L Gucma (Eds), Berlin,

pp 97–111

Shigley JE & Mischke CR (2001) Mechanical Engineering Design. 6th Edition

McGraw Hill. Inc., New York

Song BF (1997) A technique for computing failure probability of structure using

importance sampling. Computers & Structures, 72, pp 659–665

Spaethe G (1992) Die Sicherheit tragender Baukonstruktionen, 2. Neubearbeitete

Auflage, Wien, Springer Verlag

Spaethe G (1992) Die Sicherheit tragender Baukonstruktionen, 2. Neubearbeitete

Auflage, Wien, Springer Verlag

Steel RD & Torrie JH (1991) Prinsip dan Prosedur Statistika, Edition 2, Gramedia

Pustaka Utama, Jakarta, 1991

Storm R (1988) Wahrscheinlichkeitsrechnung, mathematische Statistik und

statistische Qualitätskontrolle. VEB Fachbuchverlag Leipzig

Strauss A & Bergmeister K (2005) Safety concepts of new and existing structures.

Vortrag zur fib Commission 2 Sitzung “Safety and performance concepts” am

23. Mai 2005 in Budapest

Thacker BH, Riha DS, Fitch SHK, Huyse LJ & Pleming JB (2006) Probabilistic

engineering analysis using the NESSUS software. Structural Safety 28,

pp 83–107

Tichý M (1976) Probleme der Zuverlässigkeit in der Theorie von Tragwerken.

Vorträge zum Problemseminar: Zuverlässigkeit tragender Konstruktionen.

Weiterbildungszentrum Festkörpermechanik, Konstruktion und rationeller

Werkstoffeinsatz. Technische Universität Dresden – Sektion Grundlagen des

Maschinenwesens. Heft 3/76

Tonon F & Tonon L (2006) Renovation of the 17th-Century Ponte Lungo bridge

in Chioggia, Italy. Journal of Bridge Engineering, 11 (1), January-February

2006, pp 13–20

Tvedt L (1988) Second-order reliability by an exact integral. In: Proceedings of

the 2nd IFIP Working Conference on Reliability and Optimization on

structural Systems, P Thoft-Christensen (Ed), Springer, pp 377–84

Tvedt L (2006) Proban – probabilistic analysis. Structural Safety, 28, pp 150–163

Val DV & Stewart MG (2002) Safety factors for assessment of existing structures.

Journal of Structural Engineering, 128 (2), pp 258–265

Van der Waerden BL (1957) Mathematische Statistik. Springer Verlag, Berlin

Vanmarcke E (1983) Random fields: Analysis and synthesis. MIT Press

References 331

Visodic JP (1948) Design Stress Factors. Vol. 55, May 1948, ASME International,

New York

Weibull W (1951) A Statistical Distribution of wide Applicability, J. App. Mech.,

Vol. 18, 1951, pp 253

Weiland S (2003) Das Antwort-Flächen-Verfahren. 1st Dresdner Probabilistik-

Symposium – Sicherheit und Risiko im Bauwesen. D Proske (Ed), TU

Dresden, 2nd Edition 2006

Wenzel F – (Ed) (1997) Mauerwerk – Untersuchen und Instandsetzen durch

Injizieren, Vernadeln und Vorspannen. Erhalten historisch bedeutsamer

Bauwerke. Empfehlungen für die Praxis. Sonderforschungsbereich 315,

Universität Karlsruhe

Wiese H, Curbach M, Al-Jamous A, Eckfeldt L & Proske D (2005) Vergleich des

ETV Beton und DIN 1045-1. Beton- und Stahlbetonbau 100, Heft 9,

pp 784–794

Wu Y-T, Shin Y, Sues RH & Cesare MA (2006) Probabilistic function evaluation

system (ProFES) for reliability-based design. Structural Safety, 28, pp 164–195

Yuan X-X & Pandey MD (2006) Analysis of approximations for multinormal

integration in system reliability computation. Structural Safety, 28 (4),

pp 361–377

Zhao Y-G & Ono T (1999a) New approximations for SORM: Part 1. Journal of

Engineering Mechanics, ASCE 125 1, pp 79–85

Zhao Y-G & Ono T (1999b) New approximations for SORM: Part 2. Journal of

Engineering Mechanics, ASCE 125 1, pp 86–93

Zhao Y-G & Ono T (1999c) A general procedure for first/second-order reliability

method (FORM/SORM). Structural Safety, 21 (2), pp 97–112

8 Examples

8.1 Introduction

Only a few examples can be found in literature dealing with the probabilis-

tic safety assessment of historical arch bridges. Whereas in other technical

fields such as car design, turbine design, or design of new structures, prob-

abilistic techniques are now common, especially in fields with highly

nonlinear material behaviour such as masonry, and difficult numerical ex-

pression of the overall load-bearing capacity such as found for arch

bridges, the success of probabilistic methods was so far limited. However,

the number of examples rises and a few example are mentioned here.

constructed in 1948 with a span of 60 m by probabilistic computation.

First, they investigated the safety index using only existing data, and re-

–5 –6

ceived 3.78 (7.93 × 10 ) and 4.44 (4.41 × 10 ). After extraction of updated

material properties, the safety index for different cross sections of the arch

–9 –11

increased to 5.80 (3.27 × 10 ) and 6.68 (1.17 × 10 ). Please note the

strong increase of the safety index by only considering more realistic mate-

rial parameters.

Casas (1999) has published probabilistic computations for several his-

torical arch bridges. He has used two different types of models: a linear-

elastic model that describes failure as the origin of a new hinge and a

nonlinear model that considers failure either by overload of the compression

strength of the masonry or by development of a four-hinge mechanism.

The most simple model uses the formulae

D. Proske, P. van Gelder, Safety of Historical Stone Arch Bridges, DOI 10.1007/978-3-540-77618-5_8,

© Springer-Verlag Berlin Heidelberg 2009

334 8 Examples

N M (8-1)

g=H− − 2⋅

fc ⋅ B N

with

• H as arch thickness described by a Gauss distribution with a coefficient

of variation of 10%

• B as arch width described by a Gauss distribution with a coefficient of

variation of 5%

• fc as compression strength of the masonry described by a Gauss distribu-

tion with a coefficient of variation of 20%

• M as a bending moment described by a Gumbel distribution with a coef-

ficient of variation of 10% for the live load and a Gauss distribution

with a coefficient of variation of 5% for the dead load

• N as normal force described by a Gumbel distribution with a coefficient

of variation of 10% for the live load and a Gauss distribution with a co-

efficient of variation of 5% for the dead load

Based on these input data and using the linear-elastic model Casas

(1999) computed for the Magarola Bridge the following safety indexes

(Table 8-1).

Table 8-1. Safety indexes for the Magarola Bridge for the linear-elastic model

(Casas 1999)

Cross section at No correlation r = 0 Strong correlation r = 0.9

Springing 1.54 1.64

Quarter point 4.81 4.82

Crown 1.11 1.14

The results in Table 8-1 give rather low values and indicate an insufficient

safety. The consideration of correlations decreases the uncertainty of the

input data and therefore usually increases the safety index, but in this case

not sufficiently. However, if the mechanical model is changed from type 1 to

type 2 using nonlinear failure considerations, then the safety index improves

strongly. Casas (1999) obtained safety indexes for the bridge under live

load between 10.6 and 14.3 and without live load between 11.6 and 15.5.

For two further bridges, the Jerge Arch Bridge and the San Rafael

bridge, the results are shown in Tables 8-2 and 8-3.

Table 8-2. Safety indexes for the Jerge arch bridge (Casas 1999)

Cross section at No correlation Average correlation Strong correlation

Springing 3.64 3.71 4.26

Quarter point 3.60 3.61 4.32

Crown 3.85 3.84 4.75

8.2 Examples in Literature 335

Table 8-3. Safety indexes for the San Rafael arch bridge (Casas 1999)

Cross section at No correlation

Springing 5.49

Quarter point 7.99

Crown 7.39

In Table 8-4, all results from Casas (1999) are summarized. In general,

it can be seen that linear-elastic models often yield to unacceptable safety

indexes. However, for the computation, the current traffic model by the

Eurocode 1 was assumed and it is understandable that historical bridges

were not designed for such loads. Therefore, cases where the bridges do

not meet the safety requirements in terms of a safety index are understand-

able. However, nonlinear models yield to acceptable safety indexes in all

cases. This confirms that linear-elastic models underestimate the ultimate

load of arch bridges significantly. Obviously, the historical design of the

bridges was based on linear-elastic models, and therefore the bridges were

permanently over designed. However, using more advanced models can

extend the usage of such bridges to the much higher load that we face

today.

Table 8-4. Summary of the probabilistic results from Casas (1999) in terms of the

safety index

Bridge Material Span in m β linear-elastic β nonlinear

Magarola Brick 20.0 1.6 13.0

Jerte Granite stone 22.5 3.5 6.0

San Rafael Concrete 24.0 5.5 17.6

Duenas Limestone 15.0 0.24 10.0

Quintana Concrete 19.0 7.0

Torquemada Concrete 40.0 4.5

Schueremans and Van Gemert (2001, 2004) have heavily investigated the

safety of arches in terms of safety factors and probabilistic measures. In

one example, they give the static safety factor of 2.39 and the geometrical

safety factor of 1.23 when using average material parameters.

Quan (2004) presents a probabilistic computation of a reinforced con-

crete arch. He receives safety indexes between 3.53 and 4.31 and considers

the values as too low. Quan (2004) has used an independently developed

program.

336 8 Examples

arch bridge. The mechanical model was done in the finite element program

ANSYS, however no probabilistic results are included in the publication.

Busch (1998) and Busch and Zumpe (1995) have carried out an intensive

probabilistic investigation of the Marien Bridge in Dresden, Germany.

Möller et al. (1998) and Möller et al. (2002) have also used probabilistic

investigations of historical arch bridges and vaults. Ng and Fairfield (2002)

have carried out a Monte Carlo Simulation to investigate the safety of an

arch bridge in probabilistic terms. Recently, Brencich et al. (2007) carried

out probabilistic computations. Tschötschel (1989) has carried out inten-

sive probabilistic computations of masonry, however he has not dealt with

arch bridges.

In the next section, probabilistic computations of arch bridges carried

out by the authors will be discussed.

8.3.1 Bridge 1

a six-arch bridge with a span of approximately 25 m. It was constructed

between 1872 and 1875, and it is made of regular coursed ashlar stone

work. The material is red main sandstone, a high-quality coloured sand-

stone with a uniaxial compression strength of up to 140 MPa. Toward the

end of World War II, one pier was blasted. This pier was rebuilt in 1945

and 1946 using concrete. During the reconstruction phase, the so-called

explosion chambers were built inside the pier.

Figure 8-1 shows a view of the site nowadays. A detailed description of

the bridge can be found in Proske (2003). Attention has been drawn to this

bridge because it was hit by ships in 1999.

The investigation for the bridge was split into three steps. First, the

bridge regions most severely stressed during the impact were detected by

simple numerical calculations and then selected for drilling. Since Bridge 1

consists of different materials due to the partial replacement of blasted

piers at the end of World War II, the transfer of results from one structural

element to the same element in another position was not possible. Moreo-

ver, Bridge 1 has a unique foundation. With due regard to these specific

factors, 26 drillings with a length of up to 15 m were planned and carried

out on this bridge. The drillings had an overall length of 150 m: 90 m in

masonry and 60 m in concrete.

8.3 Own Examples 337

that was used for material testing. More than 500 material tests were car-

ried out, including compressive and tensile strength tests of the sandstone,

mortar, and concrete; Young’s modulus tests; height and width measure-

ments of the sandstone; density measurements; and measurements of the

shear strength of the masonry. With the data from the tests, it was possible

to describe the material input parameters in terms of random distributions.

The choice of the distribution type has been discussed intensively in

Proske (2003). Several statistical techniques have been used to determine

the type of statistical distribution for the investigated material properties.

In addition, the strength of the concrete compressive strength samples was

multimodal. Therefore, this distribution has been decomposed into original

distributions. This statistical effect could also be identified visually on the

testing specimen and historically with documents from the reconstruction

after the war. The statistical properties of the input variables for the nu-

merical model are given in Table 8-5.

Parameter Distribution xma sb Unit

Sandstone compressive strength Lognormal 75.40 21.30 MPa

Concrete compressive strength Lognormal 47.90 22.28 MPa

Sandstone splitting strength Lognormal 4.72 1.30 MPa

Concrete tensile stress Lognormal 1.15 0.69 MPa

Young’s modulus sandstone Lognormal 28,534 7,079.6 MPa

Young’s modulus concrete Lognormal 22,552 8,682.1 MPa

Density sandstone Normal 2.27 0.15 kg/dm3

338 8 Examples

Density concrete Normal 2.26 0.10 kg/dm3

Mortar compressive strength Lognormal 11.00 7.25 MPa

Ship impact force (frontal) Lognormal 2.04 1.5 MN

Ship impact force (lateral) Lognormal 0.61 0.385 MN

Sandstone height Normal 0.7 0.13 m

Sandstone width Lognormal 0.8 0.08 m

Mortar joint height Lognormal 0.037 0.048 m

Impact height Normal 3.0 0.5 m

a

xm – empirical mean, b s – empirical standard deviation

In the second step, the mechanical behaviour of the bridge under traffic

loads and under ship impact has been modelled with a finite element pro-

gram (ANSYS). The model of Bridge 1 was particularly complex due to

the inhomogeneous structural system. Figure 8-2 permits a view inside the

bridge. Figure 8-3 shows the finite element model using area and volumes

(unmeshed), and Fig. 8-4 shows the principal compression stress inside a

hit pier. The typical element size was about 0.5 m, but smaller elements

were used in some regions. Cracks observed on the bridge, and caused by

dead and live loads, could be approved with the used models.

To achieve such results, a realistic numerical model for the description

of the load-bearing behaviour of natural stone masonry had to be incorpo-

rated into the finite element program. There exist several different tech-

niques to describe the load-bearing behaviour of natural stone structures

under normal and shear forces. The models from Mann, Hilsdorf, Berndt

and Sabha for one layer walls, and the models from Warnecke and Eger-

mann for several layer walls (Warnecke et al. 1995) have been investigated

(Proske 2003). The von Berndt (1996) model was chosen after intensive

numerical investigations to describe the load-bearing behaviour of the nat-

ural stone piers of the bridges. This model is valid for normal forces and

shear forces and therefore able to describe the load conditions under im-

pact. Also, this model has proven to reach acceptable results in the com-

parison of the load-bearing behaviour with a wide range of experimental

data. In addition, the implementation into the finite element program was

convenient.

8.3 Own Examples 339

340 8 Examples

Fig. 8-4. Principal compression stress in longitudinal section of frontal hit pier of

Bridge 1

against Bridge 1 on an IBM workstation with a power II processor took

about one hour. Of course, simple linear-elastic static and dynamic models

for the bridge have also been used to check the results of the sophisticated

models.

To incorporate the random variables in the third step, a probabilistic cal-

culation was done using the FORM and SORM methods. After the FORM

(Spaethe 1992) calculation, different SORM methods were applied to im-

prove the quality of the first. Importance sampling has also been used to

back the results (Spaethe 1992). All the methods used gave results that

were comparable from an engineering point of view.

The criteria for the probabilistic calculation included the results of the

dynamic finite element calculation. Therefore, the so-called limit state

function was not available in an analytically closed form. One way to ob-

tain results with the probabilistic calculation with a known limit state func-

tion that is not analytically closed is the application of the response surface

methodology; for example, Rajashekhar and Ellingwood (1993). This pro-

cedure was included in the finite element program ANSYS (Curbach and

Proske 1998). The FORM and SORM techniques were also incorporated

into the finite element program ANSYS using the customized capabilities

of the program. For that purpose, the techniques had to be provided as

FORTRAN subroutines and could then be compiled and linked into the

program.

8.3 Own Examples 341

8.3.2 Bridge 2

Bridge 2 has been chosen for comparison reasons with Bridge 1. Bridge 2

was built in 1893 and consists of a steel frame superstructure with natural

stone piers. Parts of the bridge were destroyed during World War II. The

static system of Bridge 2 is a four-field beam with a span of approximately

39 m. Due to the different static systems of Bridges 1 and 2, Bridge 2 will

show a different behaviour under impact. In contrast to the excellent natu-

ral stone material of Bridge 1, the material in Bridge 2 has a lower strength

(Table 8-6). Figure 8-5 shows a view of the site nowadays. A detailed de-

scription of the bridge can also be found in Proske (2003). This bridge was

chosen to represent typical historical German bridges with steel superstruc-

tures and masonry piers over inland waterways.

Parameter Distribution xma sb Unit

Natural stone compressive strength Normal 21.2 2.4 MPa

Natural stone splitting strength Normal 0.38 0.094 MPa

(lognormal)

Mortar compressive strength Normal 15.5 3.58 MPa

Ship impact force (frontal) Lognormal 2.04 1.5 MN

Ship impact force (protection) Lognormal 0.046 0.8368 MN

Ship impact force (lateral) Lognormal 0.61 0.385 MN

Impact height Normal 3.0 0.5 m

Normal load Normal 0.242 0.0242 MPa

a b

xm – empirical mean, s – empirical standard deviation

342 8 Examples

First, the maximum possible impact forces were investigated. In the next

evaluation step, the probabilistic investigation was accomplished. The re-

sults of the probabilistic investigation of Bridges 1 and 2 are shown in

Table 8-7. Several structural solutions to increase the load-bearing capa-

city of the bridges under ship impact were also investigated. They are

visualized in Fig. 8-6. The results are shown as the probability of failure,

either per impact P(V|A) or per year P(V∩A). The value per year also in-

cludes the probability of the ship impact event. To show that the models of

the bridges are comparable, the probability of failure under dead and live

load conditions were also evaluated. Lines 6 and 14 from Table 8-7 show

approximately the same value. Only these two lines refer to the failure of

the piers under normal stress, all other lines refer to the shear failure of the

piers or the arch. To allow a better comparison of both bridges, Table 8-8

summarizes the major properties of Bridges 1 and 2. The maximum per-

-6

mitted probability of failure per year is about 1.3 × 10 (E DIN 1055-100).

Due to unsatisfactory results (see Table 8-7), the description of safety in

terms of risk has been extended. However, the risk assessment is not dis-

cussed here.

Table 8-7. Probability of failure for different structural versions

Pre-stressed

Pre-stressed

Note: For both bridges, there exist different probabilities of impact – row VIII and IX for Bridge 2 and row X and XI for Bridge 1.

To compare both bridges, Bridge 1 has in row VIII and IX the same probabilities of impact applied as for Bridge 2. Note: The

probabilities are given as multiples of 10–6

a g

The pier has been found with a 3 m crack Pre-stressing of the pier with no-bond tendons (2 × 2 MN and 2 × 4 MN, respec-

b

Assumption of closing the crack tively)

8.3 Own Examples

c h

Size of the pier increased by factor 2.3 Reinforced concrete replacement type piles (2 × 3 ∅ 1,5 m) inside the pier and

d

Hypothetical material with higher tensile strength closing the explosion chamber

e i

Explosion chamber was found inside the piers Use of threaded rods (GEWI) inside the piers (2 × 4) and closing explosion

343

f

Closing of the explosion chamber chamber

f

Not considering an impact

344 8 Examples

Bridge 1 (top) and Bridge 2 (bottom)

8.3 Own Examples 345

Bridge 1 Bridge 2

Probability of failure per impact 0.023 (1.0)a 0.15 (6.5)a

Probability of failure under dead and 2.030·10–4 (1.0)a 2.400·10–4 (1.2)a

live loads

Area of the pier in m2 48 (1.9)a 25 (1.0)a

Dead load in MN 37 (7.4)a 5 (1.0)a

Existing normal stress in MPa 0.84 (3.2)a 0.26 (1.0)a

Acceptable normal stress in MPa 25 (2.5)a 10 (1.0)a

Acceptable shear stress in MPa 0.8 (2.7)a 0.3 (1.0)a

Maximal dynamic impact force in MN 13.0 (2.9)a 4.5 (1.0)a

Quantile value of the impact force in % 99.99 97.00

a

Numbers in brackets give the ratio between the two bridges

8.3.3 Bridge 3

The third example deals with a railway bridge constructed between 1911

and 1913. The bridge consists of three arch spans built of brick masonry

with concrete backfill. In the beginning, different statical models were

developed, starting with simple beam models and increasing towards

two- and three-dimensional finite element models with nonlinear material

behaviour. Figure 8-7 shows a two-dimensional model using ATENA. Be-

cause the railroads are not symmetrically located at the bridge, a three-

dimensional model was also created (Fig. 8-8). The modelling of the arch

was done in shell elements and the shell elements were fixed to the piers.

Input data were based on material tests. The material was provided by core

drilling. A nonlinear computation on an IBM working station with 2 GB

RAM required approximately one hour. Besides the deterministic compu-

tation, the probability of failure was also computed.

346 8 Examples

Fig. 8-8. Finite element model of Bridge 3 using the program ANSYS

8.3.4 Bridge 4

Bridge 4 is a railway masonry arch bridge made of natural stones (Fig. 8-9).

The bridge has an overall length of 423 m and consists of 34 arches and 33

piers. The arches have a span of approximately 10 m. The piers are distin-

guished into regular piers and group piers. Group piers simply have a

greater width (2.8 m) in comparison to normal piers (1.5 m).

The bridge was constructed in 1871 and was refurbished in 1929–1930.

Originally the bridge was designed for two railway tracks, however only

one track is in use. In general, the bridge seems to be in good condition,

but some crack damage and deformations of the spandrel walls were

found. As a result of the investigation, the cracks were related to centrifu-

gal forces and unsymmetrical loads of the train since the bridge is located

in a curvature.

The numerical load-bearing investigation started by constructing a fixed

arch, a two-hinged arch, and a three-hinge arch beam model. For those

simple models, the acceptable stress values were exceeded. Besides the

first models, drillings were also carried out at the bridge. Based on the

results in terms of material properties, the arch beam models were im-

proved (Figs. 8-10 and 8-11) and nonlinear two-dimensional (Fig. 8-12) and

three-dimensional models were created (Fig. 8-13). Considering the back-

8.3 Own Examples 347

fill and the spandrel walls, the load-bearing capacity could be temporarily

proved. If one considers that the original suggestion was the destruction of

the bridge and the construction of a new reinforced concrete bridge, the re-

sult can be seen as a success. It was recommended to improve the backfill

and add a reinforced concrete slab, however the owner agreed only to the

concrete slab. The probability of failure was also computed.

348 8 Examples

8.3.5 Bridge 5

The fifth bridge is a thought-up bridge using average geometrical and ma-

terial parameters based on references. The bridge was then modelled using

the finite element programs ATENA (Fig. 8-14) and ANSYS (Fig. 8-15).

The probability of failure per year was then computed with 8.8 × 10 .

–6

8.3 Own Examples 349

8.3.6 Bridge 6

the arch. Drilling cores were taken from the structure, and a beam model

was developed by Bothe et al. (2004) using the software package

SOFISTIK (Fig. 8-16). A probabilistic computation was then carried out to

define a partial safety factor of the arch system.

350 8 Examples

Fig. 8-16. Beam model for Bridge 6 using the software SOFISTIK (Bothe et al.

2004)

8.3.7 Summary

methods have been introduced. Table 8-9 compares the results with other

probabilistic calculations of historical bridges based on references or cal-

culated by different authors. The authors are well aware of the influence of

different mechanical models or different input data. However, the histori-

cal bridges have existed under real world load conditions for a long time

and the results of a probabilistic calculation must reflect this fact in terms

of comparability.

Proof under dead and live loads Chosen value per Reference

Mulden Bridge Podelwitz in 1888a 591.50·10–6 Year Möller et al. (1998)

Flöha Bridge Olbernhau 0.04·10–6 Year Möller et al. (2002)

Syraltal Bridge Plauen in 1905 360.00·10–6 Year Möller et al. (2002)

Bridge 1 (Bernd model) in 1875 4.10·10–6 Year Proske (2003)

Bridge 2 in 1893 4.80·10–6 Year Proske (2003)

Bridge 5 8.8⋅10–6 Year

Marien Bridge Dresden in 1846 1,279.0·10–6 Load Busch (1998)

Bridge 1 (Mann model) in 1875 33,430.·10–6 Load Proske (2003)

Bridge 1 (Berndt model) in 1875 248.0·10–6 Load Proske (2003)

Bridge 2 constructed in 1893 203.0·10–6 Load Proske (2003)

Bridge 3 343.0·10–6 Load

Bridge 6 159.0·10–6 Load

Artificial Bridge 130.0·10–6 Year Schueremans et al. 2001

Magarola Bridge 5.48·10–2–10–12 Year Casas (1999)

Jerte Bridge 2.33·10–4–10–10 Year Casas (1999)

8.4 Further Examples 351

Proof under dead and live loads Chosen value per Reference

San Rafael Bridge 1.90·10–8–10–12 Year Casas (1999)

Duenas Bridge 4.05·10–1 –10–12 Year Casas (1999)

Quintana Bridge 1.28·10–12 Year Casas (1999)

Torquemada Bridge 3.40·10–6 Year Casas (1999)

a

Limit state of serviceability

The authors have not only investigated arch bridges using probabilistic mod-

els, but also many other historical and modern structures. To show the appli-

cation to other historical structures, two further examples are illustrated.

In the Lausitz part of Saxony, Germany, many simple stone beam bridges

built of Lausitz granite are found (Fig. 8-17). Such bridges often have an

age of more than 150 years and therefore do not comply with current struc-

tural codes. To conserve these historical structures, a proof concept was

developed that not only considers special properties of the material, but

also complies with the new code generation using partial safety factors.

For the development, several tests on real-scale stone bridges were car-

ried out. The bending tests not only included original stones but also

strengthened stones. The mass of the stones reached up to 1 tonne. Besides

tests on the entire stones, compression and splitting tensile tests were also

carried out to predict the bending tensile strength indirectly. Using the test

data, statistical data investigation was carried out and characteristic bend-

ing tensile strength values, as well as partial safety factors, were given

(Curbach et al. 2004). The characteristic bending tensile strength can be

evaluated as a function of the stone height. The concept also considers the

nondestructive tests in the partial safety factor for the stone material. If ul-

trasonic or radar tests are carried out and confirm a damage-free stone ma-

terial, then the partial safety factor can be decreased from 1.7 to 1.5.

352 8 Examples

The next example considers only a part of a bridge: the numerical investi-

gation of the anchoring chamber of the Blue Wonder Bridge in Dresden

(Fig. 8-18). The bridge experienced a major flood in 2002. Although the

bridge looks like a steel framework bridge, it is actually a suspension

bridge. The top chord of the framework functions like a suspension cable

and the framework inclusive of the roadway hangs on this element. Even

further, the frameworks are constructed like independent slabs. This was

done to design a statical determined system, which was easier to calculate

during the time of construction at the end of the 19th century. The different

slabs are separated by hinges. Since a few of the hinges do not function

very well anymore, different statical systems have to be considered nowa-

days for recomputation. Whereas the dead load is still applied to the stati-

cal determined system, traffic and temperature loads have to be applied to

a statical indeterminate system. However, all loading cases function only if

the tensile forces in the top chord can be transferred to the foundations.

This is mainly done in the anchoring chamber where a counterweight is lo-

cated. During the flood in 2002, the question come up as to which water

level the anchoring chamber can bear the water pressure and when the an-

choring chamber should at least partially be flooded to limit the maximum

water pressure. If the anchoring chamber fails completely, then the coun-

terweight goes nearly completely underwater and is put under buoyant

force. However, the load-bearing capacity of the bridge then decreases.

8.4 Further Examples 353

Since the anchoring chamber is built of major stone elements with lean

concrete and has a complicated three-dimensional body, a finite element

model was used for the computation (Fig. 8-19). The nonlinear computa-

tion of a loading case ran up to one day on a PC. The input data for the

computation were partially based on material tests using test specimen

from core drillings.

As a deterministic result, the normal load-bearing capability of the

bridge can be provided up to 7 m of Dresden water level. The computa-

tion confirmed a partial filling of the anchoring chamber during the

2002 flood with a Dresden water level of more than 9 m. In relation to

the return period of such water levels, the probability of failure for the

bridge under the flooding load and without traffic restriction during

flooding has been estimated.

354 8 Examples

Fig. 8-19. Anchoring chamber finite element model and load during flooding

8.5 Conclusion

Humans can never describe the reality in one single model since all models

simplify and abstract. This is valid for social, chemical, physical, or

mathematical models. The limits of the models follow the general laws and

limits in certain areas of science. However, in contrast to scientific think-

ing, humans behave differently under everyday life conditions. They per-

manently change the limits of model borders independent from any admin-

istrative scientific types or pedagogic rules. That means, when a model

does not fit to describe real-world behaviour, the model will be changed

independent from the amount of knowledge we have. For example, we do

not have permanent scientific studies about the traffic density of roads on

hand, but we have to decide which road we would take and we are able to

do that. Models have to be developed, and have to be effective and appli-

cable. Therefore, all the models introduced here do have their meaning.

There may be cases where a linear-elastic model fits best to the conditions.

In other cases, a full probabilistic computation may be required. Even fur-

ther, in the case of historical arch bridges, we have had a development

8.5 Conclusion 355

of models for probably more than 2,000 years and even now, sometimes

the early simple rules may be of use. Such a development can be seen all

over the engineering fields. For example, for steel bridges, often we apply

advanced finite element models, however simple beam models are still

used (Unterweger 2002).

However, all humans develop their models based on their gained

information over a lifetime and their knowledge. That means that every

human develops different models based on his genetic code, his own

experience, and history. If humans reach the same results despite that

diversity using their own models, we can assume that the models function

well and are objective. The term “objectivity” can be differently defined.

For example, objectivity is the independence of the goal of an

investigation and the chosen model. In other words, objective decisions are

conscious decisions. Another definition of objectivity is based on the length

of causal relationships.

The model for the investigation of an arch bridge should also be objec-

tive. However, as stated above, the models are related to the individual.

Engineers often have different opinions and discuss different parameters,

different models, and different assumptions. Some readers may not agree

with the presented models or statements, while others do. The final ques-

tion is then are we able to describe the safety of arch structures objectively

or will we have under all conditions some subjective parts, and if so, are

we able to transfer the subjective parts into codes or rules?

We assume that research in the field of arch bridges will continue to ex-

perience progress by the objective description of the load-bearing capacity

of arch bridges, such as that shown in Schubert (2003), Purtak et al.

(2007), ICOMOS (2001), Ciria (2006), and Schiemann (2005).

However, this is only one part. We will underestimate the subjective

part when we do not consider subjective elements of bridges. In many

books, arch bridges are only dealt with in a subjective way when discuss-

ing the beauty of such bridges (Widmer 1996, Stiglat 1997, Sprotte 1977,

Pearce and Jobson 2002, Mühlen 1969, Dietrich 1998, Cruppers 1965, and

Bonatz and Leonhardt 1953). The objective modelling of complex systems

is limited: complex systems under all conditions include properties that

can only be assessed subjectively (Proske 2008).

Arch bridges are, for many people, not only a tool to cross a river or a

valley; they are part of the human history, culture, and human effort. They

make proud, they often fit perfectly into the landscape, and they are so

long-lasting. They are probably, besides temples, the technical product

with the longest usage time, sometimes reaching 2,000 years. We rather doubt

that we can currently build bridges that can be used in the year 4,000 A.D.

356 8 Examples

In everything that man pushed by his vital instinct, builds and raises,

for me, nothing is more beautiful or more precious than bridges.

Bridges are more important than houses,

more sacred because they are more useful than temples.

They belong to everybody and they are the same for everybody,

always built in the right place

in which the major part of human necessity crosses,

more durable than all other constructions.

taken from Armaly, Blasi and Hannah (2004)

References 357

References

Armaly M, Blasi C & Hannah L (2004) Stari Most: rebuilding more than a historic

bridge in Mostar. Museum International, Blackwell Publishing, No. 224, 56

(4), pp 6–17

Berndt E (1996) Zur Druck- und Schubfestigkeit von Mauerwerk – experimentell

nachgewiesen an Strukturen aus Elbsandstein. Bautechnik 73 (1996), Heft 4,

pp 222–234

Bonatz P & Leonhardt F (1953) Brücken. Verlag Karl Robert Lange Wiesche.

Königstein im Taunus

Bothe E, Henning J, Curbach M, Bösche T & Proske D (2004) Nichtlineare Be-

rechnung alter Bogenbrücken auf der Grundlage neuer Vorschriften. Beton-

und Stahlbetonbau 99, Heft 4, pp 289–294

Brencich A, Bambarotta L & Sterpi E (2007) Stochastic distribution of compres-

sive strength: effect on the load carrying capacity of masonry arches. Proceed-

ings of the Arch 07 – 5th International Conference on Arch Bridges, PB

Lourenço, DV Oliveira & A Portela (Eds), 12–14 September 2007, Madeira,

pp 641–648

Busch P & Zumpe G (1995) Tragfähigkeit, Tragsicherheit und Tragreserven von

Bogenbrücken. 5. Dresdner Brückenbausymposium, 16.3.1995, Fakultät

Bauingenieurwesen, Technische Universität Dresden, pp 153–169

Busch P (1998) Probabilistische Analyse und Bewertung der Tragsicherheit

historischer Steinbogenbrücken: Ein Beitrag zur Angewandten

Zuverlässigkeitstheorie. Dissertation. Technische Universität Dresden

Casas JR (1999) Assessment of masonry arch bridges. Experiences from recent

case studies. Current and future trends in bridge design, construction and

maintenance in Bridge design construction and maintenance, PC Das,

DM Frangopol & AS Nowak (Eds). Thomas Telford, London, pp 415–423

Ciria (2006) Masonry and brick arch bridges (RP692), www.ciria.org

Cruppers H (1965) Die Trierer Römerbrücken. Verlag Philipp von Zabern, Mainz,

1965

Curbach M & Proske D (1998) Module zur Ermittlung der Versagenswahrschein-

lichkeit. 16. CAD-FEM Nutzermeeting. Grafing

Curbach M, Günther L & Proske D (2004) Sicherheitskonzept für Brücken aus

Granitsteindeckern. Mauerwerksbau, Heft 4, pp 138–145

Dietrich RJ (1998) Faszination Brücken – Baukunst, Technik, Geschichte. Verlag

Georg D. W. Callwey, München

ICOMOS (2001) Recommendations for the analysis, conservation and structural

restoration of architectural heritage, International Scientific Committee for

Analysis and Restoration of Structures of Architectural Heritage, Draft Ver-

sion n°11, September 2001

Modena C & Sonda D (1998) Safety evaluation and retrofitting of an r.c bridge.

Arch Bridges: History, Analysis, Assessment, Maintenance and Repair. Pro-

ceedings of the Second International Arch Bridge Conference, A Sinopoli

(Ed), Venice 6.–9. October 1998, Balkema, Rotterdam, pp 223–229

Möller B, Beer M, Graf W, Schneider R & Stransky W (1998) Zur Beurteilung

der Sicherheitsaussage stochastischer Methoden. Sicherheitsrisiken in der

358 8 Examples

Technische Universität Dresden, pp 19–41

Möller B, Graf W, Beer M & Bartzsch M (2002) Sicherheitsbeurteilung von

Naturstein-Bogenbrücken. 6. Dresdner Baustatik-Seminar. Rekonstruktion

und Revitalisierung aus statisch-konstruktiver Sicht. 18. Oktober 2002.

Lehrstuhl für Statik + Landesvereinigung der Prüfingenieure für Bautechnik

Sachsen + Ingenieurkammer Sachsen. Technische Universität Dresden, pp

69–91

Mühlen F (1969) Technische Kulturdenkmale des Wasser-, Straßen- und

Brückenbaus in Westfalen. Deutsche Kunst und Denkmalpflege Jahrgang

1969, Sonderdruck

Ng K-H & Fairfield CA (2002) Monte Carlo Simulation for Arch Bridge

Assessment. Construction and Building Materials 16, pp 271–280

Pearce M & Jobson R (2002) Bridge Builders. John Wiley & Sohns, Ltd. London

Proske D (2003) Ein Beitrag zur Risikobeurteilung von alten Brücken unter

Schiffsanprall. Ph.D. Thesis, Technische Universität Dresden

Proske D (2008) Catalogue of Risks – Natural, technical, health and social risks.

Springer, Berlin – Heidelberg

Purtak F, Geißler K & Lieberwirth P (2007) Bewertung bestehender

Natursteinbogenbrücken. Bautechnik 8, Heft 8, pp 525–543

Quan Q (2004) Modeling current traffic load, safety evaluation and SFE analysis

of four bridges in China. Risk Analysis IV, CA Brebbia (Ed). WIT-Press,

Southampton 2004

Rajashekhar MR & Ellingwood BR (1993) A new look at the response surface ap-

proach for reliability analysis. Structural Safety, 12, pp 205–220

Schiemann L (2005) Tragverhalten historischer gemauerter Bogenbrücken

(Dissertation in Bearbeitung), http://www.lt.arch.tu-muenchen.de/

Schubert P (2003) Übersicht über abgeschlossene und laufende Forschungsvorhaben

im Mauerwerksbau: Realistische Tragfähigkeitseinstufung gemauerter

Bogenbrücken mit Hilfe numerischer Berechnungen. Mauerwerk Kalender,

Jahrgang 2004, p 646

Schueremans L & Van Gemert D (2001) Assessment of existing masonry struc-

tures using probabilistic methods – state of the art and new approaches.

STRUMAS 2001 18.-20. April 2001, Fifth International Symposium on Com-

puter Methods in Structural Masonry

Schueremans L & Van Gemert D (2004) Assessing the Safety of Existing Struc-

tures: Reliability Based Assessment Framework and Applications. WTA Con-

ference Prague November 23.–24., 2004

Schueremans L (2001) Probabilistic evaluation of structural unreinforced ma-

sonry, December 2001, Dissertation, Katholieke Universiteit Leuven, Fac-

ulteit Toegepaste Wetenschappen, Departement Burgerlijke Bouwkunde,

Heverlee, Belgien,

Schueremans L, Smars P & Van Gemert D (2001) Safety of arches - a probabilis-

tic approach. 9th Canadian masonry symposium, Frederiction: University of

New Brunswick, Canada, 12 pp

Spaethe G (1992) Die Sicherheit tragender Baukonstruktionen, 2. Neubearbeitete

Auflage, Wien, Springer Verlag

References 359

Kreuzwetheim, Brückenstraße 30

Stiglat K (1997) Brücken am Weg. Frühe Brücken aus Eisen und Beton in

Deutschland und Frankreich. Ernst und Sohn, Berlin

Tschötschel M (1989) Zuverlässigkeitstheoretisches Konzept zur Bemessung von

Mauerwerkskonstruktionen. Dissertation, Technische Hochschule Leipzig

Unterweger H (2002) Globale Systemberechnung von Stahl- und Verbundbrücken

– Leistungsfähigkeit einfacher Stabmodelle, Habilitation, TU Graz

Warnecke P, Rostasy FS & Budelmann H (1995) Tragverhalten und Konsolidierung

von Wänden und Stützen aus historischem Natursteinmauerwerk.

Mauerwerkskalender 1995, Ernst & Sohn, Berlin, pp 623–685

Widmer M (1996) Eisenbahnbrücken. Schweiz, Deutschland, Österreich. Transpress

Verlagsgesellschaft, Stuttgart

Žak J, Florian A & Hradil P (2003) A contribution to the analysis of historical

structures using LHS method. Structural Studies, Repairs and Maintenance of

Heritage Architecture VIII, CA Brebbia (Ed), WIT-Press 2003, Southampton,

pp 252–256

Index

Archtec, 248

Abutments, 28 Arithmetic mean, 313

Additional-leaf masonry, 188 Assessment, 100, 265

Aging, 217 Assessment scheme, 14

Aita method, 124 Auxerre-traffic, 77

Alcántara, 3

Analysis method, 100 B

Aquaduct, 25

Arch Backfill, 25, 130, 133, 142, 238

basket, 34, 35 Basket arch, 34, 35

catenary, 34 Beam models, 125

circular, 34 Behaviour

cycloid, 34 stress-strain, 139

elliptic, 34 Binomial distribution, 323

geometry, 34 Blue wonder bridge, 75

half wave, 34 Boldness

irregular, 32 coefficient, 116

maximum length, 115 Breaking forces, 87

parable 4th order, 34 Bridge, 345, 347, 352

parabolic, 34, 35 Bridge in bridge, 239

sinus curve, 34 Bridge class

skewed, 30 road traffic, 77

thickness, 104, 111 Bridge types

Arch bridges statically, 29

computation, 102

concrete, 49 C

history, 36

medieval, 40 Catenary arch, 34

number, 50 Censored data, 314

skewed, 33 Characteristic load

steel, 47 road traffic, 69, 70

stone, 36 traffic moments, 80

terms, 20 Characteristic value, 310

wooden, 48 Chemical weathering, 226

D. Proske, P. van Gelder, Safety of Historical Stone Arch Bridges, DOI 10.1007/978-3-540-77618-5_BM 2,

© Springer-Verlag Berlin Heidelberg 2009

362 Index

Circular arch, 34 Destructive test, 202

Cleaning, 236, 241 Deterioration, 219

Coefficient of boldness, 116 Diagrams

Collapse, 129, 221 moment-axial force, 187

Commercial programs Discontinuities, 218

probabilistic computation, 295 Discrete element method, 140

Compound beam model, 133 Displaced stone, 234

Compression strength Distribution

masonry, 176 binomial, 323

mortar, 173 lognormal, 318

sandstone, 169 normal, 316

Computation Weibull, 318

historical stone arch bridges, 99 Drilling, 203

Concrete arch bridges, 49 Drying, 237

Conductivity

electrical, 208 E

Conservation, 237

Contamination, 218 Economical growth, 39

Contour scaling, 223 Effective bending stiffness, 134

Copula, 294 Effective height, 128

Core, 204 Effective width, 148

Cornigliano Bridge, 27 Egyptians, 36

Correlation, 294, 316 Electrical conductivity, 208

Corrupted data, 314 Electric resistivity, 200

Coverage, 129 Elliptic arch, 35

Crack, 219, 229, 230, 231 Empirical rules, 100

diagonal, 234 Endoscopy, 201

longitudinal, 231 Etruscans, 36

transversal, 232 Evaluation, 17

Crown, 104 Exfoliation, 223

Curvature, 23, 33 Experimental test, 208

Cycloid arch, 34

F

D

Fabricius bridge, 37

Damages, 14, 217, 218, 219 Failure, 221

spandrel walls, 234 types, 101

Data, 314 Failure surface, 138, 139

Dead load, 92 Fatigue, 220

Debris flow impact, 221 FILEV method, 121

Deformation, 218, 220, 227 Finite element method, 136

Degradation, 220 First order reliability method, 270, 340

Design work time, 217 Flaking, 223

Index 363

Flooding, 221 inspection, 236

FORM, 270, 340 Interventions

Front wall, 25 structural, 18

Investigation, 199

G

J

Geometry

arch, 33 Joint pattern

Global safety factor, 302 skewed bridge, 33

Greek, 36

Grouting, 237, 246 K

Growth

economical, 39 Kurtosis, 313

H L

Harvey method, 123 Load

Height breaking forces, 87

effective, 128 dead load, 92

Hinge, 132 impact, 89

Historical rules, 100 initial drive forces, 86

Historic research, 200 railroad traffic, 82

History road traffic, 67

arch bridges, 36 settlements, 89

Hypersphere division method, 281 snow, 92

temperature, 89

I wind, 88

Load border system, 85

ICOMOS, 14 Load test, 141

Impact-echo, 206 Lognormal distribution, 318

Impact forces, 89 Lorry classes, 73

Importance Sampling, 340 Loss of material, 218

Income, 38

Increase of width, 252 M

Infrared thermography, 201

Initial drive forces, 86 Maintenance, 56, 235, 236

Injection, 238, 246 Maintenance cost, 56

Inspection, 236 Martinez method, 124

Inspection concepts, 14 Masonry, 165, 338

Inspection interval, 236 additional-leaf, 188

364 Index

model Monitoring, 201

Berndt, 181 Monte Carlo Simulation, 285

DIN 1053, 178 crude, 285

DIN 1053–100, 177 variance reduced, 286

Ebner, 185 Mortar, 166, 172

empirical exponential, 179 historical, 173

Francis, 184 Mortar refurbishement, 243

Hilsdorf, 179 Movement, 219

Khoo & Hendry, 184 Multimodal data, 314

Mann, 180 Multi ring arch, 26

Ohler, 183

Sabha, 183 N

Schnackers, 185

Stiglat, 184 Nailing, 237

Masonry strength, 165 NDT, 201

MDT, 202 Non-destructive test, 201, 205

Mechanical cleaning, 237 Non-metallic reinforcement, 249

Method Normal distribution, 316

Aita, 124 Number of arch bridges, 50

Breitung, 274

Cai & Elishakoff, 280 O

discrete element, 140

FILEV, 121 Observation concepts, 20

finite element, 136 Outlier, 314

Harvey, 123 Overload, 219

Jaeger & Bakht, 321

Köylüoglu & Nielsen, 277 P

Leicester, 319

Martinez, 124 Parable 4th order arch, 34

MEXE, 119 Parabolic arch, 34, 35

Öferbeck, 320 Parapet, 251

Pauser, 124 Partial safety factor, 303, 304, 305

Purtak, 123 Pauser method, 124

response surface, 281 Per capita income

Tvedt, 276 historical, 38

MEXE method, 119 Perronet, 42

Minor-destructive tests, 202 Photogrammetry, 200, 209

Model Pier, 114

beam, 125 Piles, 238

compound beam, 133 Pillar, 29

Egermann, 188 Plasticity, 132

single beam, 125 Plasticity theory, 132

Warnecke, 188 Pont du Gard, 3

Index 365

Principal compression stress, 340 DIN-report, 69

Probabilistic calculation, 340 Eurocode, 69

Probabilistic safety concept, 269 Roadway quality, 73

Probability of failure, 14, 266, 267, Romans, 37, 39

270, 343, 353 Rules

Probability function, 312 empirical, 100

Profile bars, 249 historical, 100

Proof equations, 191

Proof loading, 201 S

Purtak method, 123

Safety, 15, 139, 265, 266

Q Safety assessment, 265

Safety concept, 182, 268

Quasi-random number, 287 probabilistic, 269

semiprobabilistic, 301

R Safety factor, 139

global, 302

Radar, 201, 207 partial, 303, 304, 305

Radiography, 200 Safety index, 270, 333, 334, 335

Railroad traffic, 82 combination, 288

load border system, 85 goal values, 297

Railway load patterns, 85 Salt attack, 224

Random field, 294 Sealing, 237

Recalibration classes Second order reliability method,

road traffic, 71 274, 340

Refurbishement, 243 Semi-destructive test, 205

Refurbishment, 236, 240 Semiprobabilistic safety concept,

Regensburg bridge, 40 301

Reinforced concrete slab, 238, 244 Separation, 219

Reinforcement, 245 Serviceability, 192

non-metallic, 250 Settlement, 89, 232

Repair, 217, 234, 236 Shear strength, 190

Repoint, 238 Shotcrete, 238, 239, 250

Response surface, 283 Simulation

Response Surface Method, 281 road traffic, 76

Restoration, 251 Single beam model, 125

Rise, 22, 102 Sinus curve arch, 34

Risk based inspection, 20 Skewed arch, 30

Road traffic Skewed masonry arch bridges, 33

bridge class, 77 Skewness, 32, 313

load influencing factors, 72 Snow loading, 92

recalibration classes, 71 SORM, 274, 340

Road traffic loads, 67 Spalling, 223, 224

366 Index

Spandrel wall, 25, 130, 142 Thickness

damages, 233 arch, 104, 111

Spare vault, 25 foundation, 112

Springing, 23, 27, 28, 111 Threaded rods, 248

Standard deviation, 313 Tomography, 208

Steel arch bridges, 47 Traffic, 72

Stiffness Trajan, 38

effective, 134 Transverse direction, 148

Stochastic structural analysis, 294 Types of failure, 101

Stone, 166, 167 Typology

Stone arch bridge, 21, 32 stone arch bridges, 32

elements, 22

Stone refurbishment, 241 U

Strength

masonry, 165 UIC 71, 83

shear, 190 Ultrasonic, 201, 206, 207

Strengthening, 235, 237

Stress distribution, 187 V

Stress-strain behaviour, 139

Stress-strain relationship, 171, 186 Vault, 21

Structural interventions, 18 Vehicle weight, 76

Sumerian, 36 distribution function, 74

Supplement, 236 Vibration test, 201

Surface Visual inspection, 200

failure, 138, 139

W

T

Weathering, 222

Technique biological, 226

evaluation, 200 chemical, 225

investigation, 199 mechanical, 226

repair, 236 physical, 226

Temperature loading, 89 Weibull distribution, 318

Term, 20 Width

arch, 20 effective, 148

vault, 21 increase, 252

Test Wind loading, 88

destructive, 202 Wooden arch bridges, 48

experimental, 208 Work time, 217

minor-destructive, 202

non-destructive, 201, 205 X

semi-destructive, 205

Testing, 141 X-ray microscope, 204

## Molto più che documenti.

Scopri tutto ciò che Scribd ha da offrire, inclusi libri e audiolibri dei maggiori editori.

Annulla in qualsiasi momento.